future of journalism essay

Try searching for

• Misinformation
• Subscriptions
• Fact-checking

AI and journalism: What's next?

Illustration generated by the Midjourney 5.2 text-to-image model, using the prompt “An abstract image representing the uncertain future of digital journalism in the age of artificial intelligence.”

Innovation in journalism is back. Following a peak of activity in the mid 2010s, the idea of fundamentally reinventing how news might be produced and consumed had gradually become less fashionable, giving way to incrementalism, shallow rhetoric and in some cases even unapologetic ‘innovation exhaustion.’ No longer. The public release of ChatGPT in late November of 2022 demonstrated capabilities with such obvious and profound potential impact for journalism that AI-driven innovation is now the urgent focus of the senior leadership teams in almost every newsroom. The entire news industry is asking itself ‘what’s next’?

For many people in journalism the first half of 2023 was a time for asking questions and learning the basics of AI. What can ChatGPT actually do? What is generative AI? What is a language model? What is a ‘prompt’? How dependable are these tools? What kind of skills are required to use them? How fast is this technology improving? What are the risks? How much of all this is just hype? 

Many newsrooms went further, providing their employees and audiences with statements or guidelines describing how they intended to approach the use of generative AI in their workflows and news products. Some even began publishing a few experimental articles written by ChatGPT. Very few, however, have yet taken specific steps to pragmatically and routinely apply these technologies in their newsrooms. Change is in the air, but specific initiatives are harder to find.

Over the past six months I have had the privilege of spending significant time discussing AI with the senior leadership of more than 40 news organisations, ranging from scrappy digitally-native newsrooms in Asia, the Middle East, and Latin America to many of the largest global news providers based in the US, UK, and Europe. 

This access came as a result of my familiarity with AI-driven innovation in news, acquired from a combination of applied product innovation at large media companies in the US and UK, and from a small parallel academic career exploring the first principles of computation applied to news production and consumption. 

My journey with many of the questions now facing news organisations began in Silicon Valley more than a decade ago, and my exploration of the GPT family of language models began in 2019 at the BBC. My career focus — a relative niche until the arrival of ChatGPT — was suddenly of great interest to many people in the news industry.

This article is based on a distillation of what I have learned from these conversations. It assumes a knowledge of generative AI’s general capabilities and potential, and examines some of the ways that large news organisations are thinking about its strategic and practical implications for their newsrooms. My intent here is to help advance the conversation beyond awareness and towards specific initiatives that can help move newsrooms forward in preparing for an uncertain AI-mediated future. 

Any point of view regarding the application of AI to news is somewhat speculative, especially given the remarkable pace of advancement in AI functionality. Nonetheless, I believe that some clarity may be emerging.

Coming up with a list of things that your newsroom might be able to do using ChatGPT is fairly easy (Summarisation! Rewriting in a simpler style!). It is much harder to clearly identify exactly what it is that you are trying to achieve with generative AI, why you are trying to achieve it, how you might plausibly achieve it at scale in routine and professionally managed operations, and whether that achievement will even continue to be relevant as AI fundamentally alters the competitive landscape in the coming months and years.

To frame this discussion I will begin with an overview of potential strategies for using AI in news, before turning to options for the practical deployment of AI in newsrooms, and the infrastructural and organisational requirements needed to support those options. At the end of the piece I will offer a point of view for changes that we might anticipate for the news industry in the longer term.

1. Applying generative AI to news: product and editorial strategies

Efficiency-focused strategies

The most obvious opportunities for applying generative AI to news are in bringing new efficiencies to specific and familiar steps within the existing news production workflows supporting an organisation’s existing news products. This ‘more-efficient-production-of-existing-products’ strategy is attractive in its simplicity, but its benefits will almost certainly be short-lived because it assumes that the existing media environment will continue in roughly its existing form. 

It is increasingly possible that the competitive environment, product offerings, production workflows and business models of news organisations will change, possibly radically, as use of generative AI becomes ubiquitous and as AI-based media products appear. 

There are early indications of this in the nascent generative search experiences offered by Google and Microsoft, in the user control of consumption experiences offered by well-funded news aggregators such as Artifact (launched by the founders of Instagram), in the declared intention to build a foundation language model around ‘X’ (or Twitter 2.0), and even in the early behaviour of tens of millions of early adopters of generative AI tools like ChatGPT and Midjourney, who have quickly learned how to use traditional media essentially as raw material for their own self-directed consumption experiences. 

A ‘more-efficient-production-of-existing-products’ strategy is clearly a reasonable place to start, but it does not fundamentally compete with new AI-enabled experiences and therefore may not remain sufficient for long.

Product expansion strategies

A more ambitious AI strategy for news organisations lies in reimagining and expanding the scope and scale of an organisation’s news products in ways that only become feasible using generative AI. This ‘new-products-for-new-audiences’ strategy will likely be more enduring than an efficiency-only approach because it seeks to actively accommodate the expansion of audience choice that generative AI enables.

In essence this approach assumes that if generative AI is going to provide audiences with more control over their news consumption experiences (as pioneered by the Artifact app), then news organisations might as well offer that option directly. If news consumers can now control how they consume articles, then shouldn’t they be able to do that on a news publisher’s app or website? 

The advantage of this strategy is that it offers a way to compete within an emerging AI-mediated news ecosystem while continuing to exploit the advantages that news organisations already possess — brand, trust in editorial processes, habit, etcetera. 

Furthermore, that competition need not merely be aimed at existing targeted audiences but might now also be extended to entirely new audiences via new journalistic products that would have been editorially impossible without generative AI. Even small newsrooms now have the option of offering comprehensive multi-media news products to diverse audiences with radically different consumption behaviours and preferences.

Differentiation strategies

A more enduring strategy for news in the age of AI will necessarily be centred on differentiation and competitive advantage, offering exclusive news products that remain uniquely valuable to audiences even as the information ecosystem changes. 

This ‘unique products’ strategy is challenging because it will be audiences that decide on the relative uniqueness of a newsroom’s products, not newsrooms. Pursuit of differentiation must therefore be clear-eyed, unsentimental and data-tested – magical thinking about vague specialness requiring little effort to implement is unlikely to be helpful here. A differentiation strategy might still potentially fit well with some of the values and brand attributes of traditional journalism, especially if the information ecosystem begins to significantly degrade under the onslaught of AI-generated content. 

Some examples of familiar forms might include: proprietary news gathering using special access or special skills; audience trust gained from well-demonstrated verification and validation processes; richly contextual content produced from archives or from deep familiarity with subject matter; or perhaps a commitment to systematic coverage of particular subject domains using narrowly customised stories, such as local coverage or climate reporting. 

All these opportunities, however, will likely need to be comprehensively optimised using generative AI to remain competitive within an AI-mediated information ecosystem, even if their core value is differentiated. Developing forms of differentiation will likely be very challenging for many news organisations, especially for those whose current product is largely built on packaging commodity information, however there may eventually be no alternative.

Techno-editorial strategies

A similar but even more ambitious potential strategy, probably available only to a few well-resourced news organisations or well-funded start-ups, is to seek to develop proprietary or competitively defensible information technologies and services centred on news – products that might resemble specialised intelligence tools more than traditional news products. 

This option, which we might call a ‘techno-editorial product’ strategy, would likely reverse the relationship between the product and editorial functions that currently exists in most news organisations, requiring editorial operations to support product and business development rather than the other way around.

 This kind of strategy might include provision of highly technical solutions such the systematic extraction of news events (and, critically, of news stories) from enormous streams of natural language text and speech, the creation and maintenance of proprietary datasets in story form, the systematic certification of that news, new tools for exploring and contextualising complex news, the extreme customization of news and of the experience of consuming news, etc. 

Such processes and tools might resemble those used by government intelligence organisations, platforms like search engines, or processes and tools already offered by news organisations targeting financial professionals, but with much broader coverage and audience appeal. 

Such a strategy would likely be research-led, would require substantial capital investment and needs to be justified by expected returns from high-value subscriptions. Such ‘techno-editorial’ businesses may not be well suited for reach-based news products but might nevertheless become significant components of an information ecosystem in which generative AI is ubiquitously deployed.

Training your own model

As news organisations look for strategic responses to generative AI, there are also a few options that are often discussed but which might be less attractive than they first seem. One of those ‘approach-with-caution’ strategies is that of pre-training a proprietary language model using a newsroom’s archive.

This is likely a bad idea for most news organisations for multiple reasons, including the difficulty in attracting the world-class machine learning talent needed to create a useful language model, the tiny size of the archives of even the largest news organisations relative to the training needs of a useful language model, and the substantial costs of periodic retraining, maintaining and operating a proprietary language model in the face of rapidly improving technology. 

Furthermore there are few clear benefits available from training a proprietary model, even if it were competitive in performance to a commercial or open-source language model. Concerns about data security, vendor lock-in or a need for niche functionality can probably be addressed in much easier ways, including contractually, via abstraction of user interfaces, using model fine-tuning or using an open-source model. 

A proprietary model built on the same transformer architecture as most current LLMs will still hallucinate, even if the entire training corpus is accurate, and a news organisation is unlikely to match the efforts underway by technology companies aimed at making language models more accurate. Training a proprietary language model may make sense for some very specific use cases in some news organisations with special information products (for example a ‘ text-plus-financial-data’ multi-modal model at Bloomberg), but even in such cases a clear-eyed, careful consideration of the costs and benefits of such a strategy is essential.

Building a chatbot for archives

Another ‘approach-with-caution’ strategy that some news organisations have pursued or considered is that of building a proprietary chatbot that enables conversational, interactive access to the organisation’s current news and archive. 

This option seems attractive given the success of ChatGPT. However, when considering a news chatbot, it is important to separately consider the interface component (i.e. interactive chat), the information component (e.g. news and its context) and the underlying technology used to deliver the experience (e.g. LLMs). 

Despite the rapid adoption of ChatGPT, text or verbal chat as an interface is still very far from broadly accepted as a way of accessing any kind of information, much less news. Audience analytics for news bulletins built for voice agents like Amazon Alexa and Google Home have been disappointing, and early attempts at LLM-enabled chatbots by media companies have delivered little traffic. 

This lack of success might be at least partly due to the relative inadequacy of news and news archives as a dataset powering such an interface, which for news organisations will be considerably smaller and more homogenous than the vast datasets that power ChatGPT or even search for that matter. 

Furthermore, from a technology perspective it is very difficult to replicate the fluid nature of communication with an LLM using the techniques currently available for constructing chatbots from archives, which typically involve some kind of embedding-based search operation combined with interpretation of the results by the language model. 

The result is often either far too many ‘I don’t know’ responses to specific queries, or references to ‘closest match’ archived articles in isolation from current context, an experience that can’t compete with internet-enabled chatbots like Bing Chat or Perplexity.ai.

Communication strategy

Regardless of where a news organisation might be in its path towards a strategy for responding to AI, there is an urgent strategic need that requires early attention – that of communicating the organisation’s approach to AI to stakeholders, funders, staff and audiences. 

The urgency of strategic communication about AI for news organisations comes from the awareness that every individual connected with the organisation already has about the abilities of these tools and their potential for dramatic disruption. Most people in the news industry are already playing with the tools, reading the reports, assessing the potential and asking the obvious questions. 

It would, of course, be ideal to patiently assess the situation, carefully devise a strategy, produce a plan for executing that strategy, and then communicate that strategy and plan – but that may not be possible to do quickly. 

A more practical communication strategy at this stage might consist of acknowledging the situation, articulating how the organisation is engaging with and learning about AI, providing clear guidelines for its early or limited use, demonstrating new approaches to innovation, signalling adaptability and generally preparing for change. A tone of optimism and excitement for the potential of AI to help raise journalism above commodity information and to make it more accessible to many more people is also appropriate.

A strategic path to continued relevance

The common objective for almost all news organisations in navigating the coming AI-induced transformation is, bluntly, continued relevance. 

News organisations, whether funded by ads, subscriptions, public funds or philanthropy, will seek to remain sufficiently valuable to enough people to ensure that those funding mechanisms continue to function. 

The challenge of the coming transition to an AI-mediated information ecosystem is therefore to navigate a path that takes advantage of near-term opportunities for efficiency, medium-term opportunities for new products and services for audiences, and longer-term opportunities for re-imagining what news can become in a transformed information ecosystem. 

This path obviously comes with considerable associated risk, specifically the risk of losing existing value and of not being able to develop new value to replace it, and may therefore require a greater tolerance of risk by leaders in news organisations. Developing that greater tolerance for risk, a tolerance perhaps closer to that of the technology companies that initiated this transition, might be the most important strategic step that a news organisation can take.

2. How to deploy generative AI in newsrooms

From strategy to projects

Any strategy for adapting a newsroom to an information ecosystem defined by generative AI is of little use without specific, practical projects that translate that strategy into useful outcomes. 

Identifying such projects is obviously difficult during this current period of exceptionally rapid change, because of the considerable risks of wasted investments, embarrassing missteps or potential damage to brands or trust. In addition, projects can only contribute meaningfully towards a strategy if they can be applied in routine, day-to-day operations, rather than merely as testable prototypes or demonstrators. 

To move forward, therefore, we need to identify categories of projects that might advance newsrooms towards an AI-ready future while minimising risks, and also identify the infrastructure requirements needed to deploy those projects routinely at scale when that time comes.

Back-end projects

A relatively low risk category of generative AI projects is purely back-end applications. These are applications with no direct audience-facing output, not even draft text, but which instead deliver their value to journalists or to businesses. These can include efficiency-focused or quality-focused tasks such as tagging, other kinds of categorisation, headline and SEO suggestions, assembly of newsletters from pre-existing copy, copy-editing, brainstorming and ideation, early research, some analysis, etcetera. 

In addition to being relatively low risk, back-end AI applications are also relatively easy to implement as they are often ‘loosely coupled’ to news production workflow and infrastructure. Such applications can sometimes be managed by stand-alone tools disconnected from the primary publishing stack of the organisation, operated by specialised staff separate from the main editorial workflow.

Language task projects

A more ambitious but still relatively low risk category of generative AI projects include those applications that produce draft text by modifying source text in some way. These applications use language models solely for ‘language tasks’ and should not introduce any information content into the draft that was not already present in the source document. 

They explicitly do not depend on the knowledge available to the model as a result of its training data, and they therefore reduce (but don’t eliminate) the risks of hallucinations, biases and other issues. 

Examples of language tasks include summarisation , simplification, stylistic re-versioning, rewriting text for particular channels (social media, topic bulletins, etcetera), script-writing for audio or video, translation... 

Language tasks can be done using any form of model access, such as the ChatGPT user interface, custom-built API-driven tools or even as new features integrated into content authoring and management systems. Language tasks can also be part of any strategy, including strategies based on efficiency, on new products or on differentiation. They are a fundamental category of journalistic task in an AI-enabled newsroom.

Knowledge task projects

A higher risk category of generative AI projects are applications that produce draft text with information content that originates in the language model itself, rather than from a source document. 

These applications perform so-called ‘knowledge tasks’, because they are doing more than merely modifying language – they are true ‘authoring’ applications of language models. 

The increased risk associated with knowledge task projects comes from the significant potential for hallucinations, simple error from training data, biases, out-of-date context and other limitations inherent in language models. 

Nonetheless, if these risks can be managed , knowledge tasks offer a substantial range of news products including provision of context for stories, explainer-like background content , different interpretations of events based on historical context, and even full articles, especially on evergreen, commodity subject matter. 

Mitigating the risks associated with knowledge tasks requires editing processes designed to detect error and inappropriate content – non-trivial tasks, as evidenced by the difficulties experienced by CNET in publishing knowledge-based content from language models. Nonetheless, there are clearly many significant journalistic opportunities available from knowledge tasks, and these opportunities will likely increase as we collectively learn more about how to manage and edit their output and as we begin to use AI to help do that. 

As with language tasks, knowledge tasks can be integrated into workflows in different ways and can contribute to different strategies. They too are a fundamental category of journalistic task in an AI-enabled newsroom.

Medium-to-medium transformation projects

A particularly ambitious category of generative AI projects for news are applications that transform information content from one information medium into another, for example from text into audio, from text into video or from text into graphical images. 

Unlike language tasks and knowledge tasks, these applications typically depend on special-purpose medium-to-medium transformation models, often used in combination with general-purpose large language models within complex workflows. 

Such special-purpose models include speech-to-text models (transcription), text-to-speech models (synthetic voices), text-to-video models (synthetic avatars, automated generation of B-roll video, etcetera), text-to-image models and others. These tools are still at an early stage. But they are developing very quickly, are widely available, and already easily match human quality in many cases. 

Furthermore, the potential of this category of applications is likely to increase with the imminent arrival of so-called ‘multi-modal’ functionality enabling richly descriptive image-to-text, video-to-text and other transformations. 

Examples of potentially high-value journalistic tasks that can be accomplished using these cross-media models include the automated or semi-automated creation of text articles from audio or video source material, the creation of audio and video news products from text articles , the transformation of text articles into graphical stories or videos, and the automated or semi-automated creation of podcasts from articles. 

These tasks are most useful for a product expansion strategy. They can often be achieved using just model vendor user interfaces and so have the potential to enable relatively low-resource newsrooms to quickly offer multimedia content at significant scale. 

Some potential barriers to implementing this category of projects include the need for an editorial producer with experience in the output medium to ensure quality, the not insignificant cost of using the specialised models, and the challenge of distributing the same story in several different media.

Listening and monitoring projects

The project categories described above all focus on novel ways of producing news products using AI, but practical projects focused on news-gathering are also viable using large language models. 

The term ‘generative AI’ has an obvious built-in bias towards the generation of media, but large language models and multi-modal models can also read, listen and soon observe at enormous scale, enabling entirely new ways of reporting news. 

The kinds of news-gathering tasks that these models can perform extend far beyond the earlier generation of ‘social listening’ tools, which usually just searched for keywords in the feeds of social media platforms. The ‘natural language understanding’ (NLU) abilities of LLMs can not only read, but interpret, evaluate, analyse, synthesise and summarise. 

Furthermore, they can do this not just with natural language text and speech, but also with structured data — as clearly demonstrated by OpenAI’s new ‘ code interpreter ’ add-on to their GPT-4 LLM, and likely soon with visual information in images and video. Projects based on NLU and ‘reporting at scale’ are already underway at newsrooms, including small newsrooms. Such projects may be most appropriate for a differentiation strategy and might enable newsrooms to build out and defend special abilities to report systematically on specific domains.

Advanced projects

Our understanding of potential applications of AI in news is still nascent, not only because of the rapid development of new functionality but also because we have just begun to explore the potential of these tools for news work. 

The projects described above are easily possible right now, with existing functionality, but there are also several ‘near frontiers’ of functionality that will likely open up significant new potential. One of these is the advent of multi-modal functionality, which is already available in limited forms in Bing Chat, MidJourney and soon GPT-4, and is likely to become increasingly powerful for applications that cross between language and visual information. 

A second near frontier is the advent of LLM ‘agents’, pioneered by small examples like AutoGPT and BabyAGI . These approaches use LLMs to deconstruct high-level tasks expressed in vague terms into small, specific, actionable tasks that are then carried out by the LLM. They offer intriguing possibilities for automating some investigative journalism, and for scaling investigative reporting into an ongoing and systematic function.

The AI strategies and specific AI projects described here illustrate that newsrooms have tangible and specific options for moving forward towards an AI-mediated information ecosystem, but these strategies and projects are insufficient. Producing professional AI-enabled news products, using AI-enabled workflows, at sufficient scale to make a difference, day-after-day and month-after-month, requires something more. It requires infrastructure.

3. Infrastructure for an AI-ready newsroom

Old-school AI infrastructure

In the decade prior to the rise of large language models, infrastructure for AI in news organisations meant something different than it does today. It meant data warehouses and data lakes, a well-structured and well-maintained metadata schema, libraries of embeddings, an expensive data science team, a large monthly AWS bill and a product roadmap focused on training small, specialised ‘machine learning’ models from scratch using small volumes of proprietary data. 

These roadmaps often included business-focused models predicting propensity to subscribe, propensity to churn or willingness-to-pay, journalist-focused models enabling a host of special functions useful for a handful of special stories, and audience-focused models such as semantic search and, of course, different kinds of recommender systems. 

This kind of AI infrastructure is still very valuable and useful, even if affordable only by a small number of elite news organisations, but it is very different from the infrastructure required to apply generative AI.

Professionalised prompt management

A fundamental and permanent requirement for applying generative AI to news work is infrastructure that enables the professional development, testing and deployment of prompts. 

AI models will be a permanent part of the future of news, and controlling those models in the service of useful journalistic work will be a central function of editorial organisations. 

That control will be exercised through prompts. Whether back-end tasks, language tasks or knowledge tasks, and whether employed as part of an efficiency strategy, a product expansion strategy or a product differentiation strategy, all applications of generative AI in newsrooms are fundamentally dependent not just on the models used to execute them, but also on the prompts used to direct those models. 

Permanent, professional mastery of prompts does not look like journalists casually cutting-and-pasting from a dozen options in a Google Doc into ChatGPT, but instead looks like a professional ‘prompt-to-publish’ pipeline that enables systematic and quality-controlled management of every aspect of prompting, outputs, editing and deployment. 

This includes prompt design, assembly, evaluation and testing, ‘certification’, metadata, storage and retrieval, versioning, iterative improvement, usage tracking, analytics, output editing, training and more. Even just prompt design can involve explicit task definition, application of system roles and ‘custom instructions’, development of few-shot examples, management of context size, prompt templating, multi-prompt staging and more. 

Just evaluating the outputs of prompts applied to stochastic models fed by diverse source documents presents a combinatorial editing challenge unlike anything previously seen in news work. All of this requires infrastructure – databases, tools, user interfaces, schemas, integration, processes, analytics, training and documentation.

Interfaces between prompts and journalistic tasks

Assuming that adequate ‘prompt-to-publish’ infrastructure is in place, a newsroom still requires an interface between this infrastructure and its journalists. 

For many journalists this is unlikely to be the raw prompt, which for useful tasks will likely be long, quite complex and probably ‘certified’ as tested and reliable according to some accepted quality control process. Instead most journalists will likely access prompts via buttons and controls that produce draft outputs for editing and refinement. 

Such interface infrastructure could consist of either stand-alone journalistic ‘co-pilot’ tools with separate experiences for prompt management and prompt use – something already seen in Google’s ‘Genesis’ interface , in various ‘News AI’ tools from start-ups and in some in-house tools developed by newsrooms. It will also require a comprehensive integration into an organisation’s existing content management infrastructure or possibly even as an entirely new form of CMS designed specifically for AI workflows. 

Maintaining control of these tools, of the buttons and functions they provide, and of the prompts behind those buttons will likely be critical for journalistic independence in an AI-mediated news ecosystem.

Infrastructure for personalised experiences

Even with a coherent strategy, a set of applications that support that strategy and a prompt management process that professionalises the execution of those applications, the extent of AI-enabled news production will still be limited by the available content management, serving and distribution infrastructure. 

This is especially true of a product expansion strategy executed using language tasks or medium-to-medium transformations. 

Producing 5, 10 or 20 different variants of every story, suitable for a wide variety of audiences, situations and consumption contexts, might be far easier than actually serving each of those variants to the right user at the right time. This is a personalisation challenge, but one that is quite different from the usual interpretation of ‘personalisation’ in most newsrooms – which tends to focus almost exclusively on personalised recommendations of one-size-fits-all content artefacts. 

This new requirement is, instead, for personalisation of the story experience, and it therefore requires infrastructure that can store, select and serve different variants of a story to different users or different user segments in different situations – in addition to continuing to recommend relevant stories. 

A simple form of experience personalisation, being pursued by several newsrooms, is to set up a separate channel or even an entirely separate brand from which to serve a newsroom’s stories in new AI-generated formats. 

A more advanced form is to offer direct user control over their consumption experience, for example in the way that the Artifact news app offers article consumption as summaries, simplified text, emoji stories, poems or as audio readings by celebrities (see illustration above). 

A more complex form of experience personalisation is to automatically adapt a user’s experience using behavioural data and contextual signals. 

Each of these approaches can open up access to a newsroom’s journalism to more people, but each also requires infrastructure that can support the management and serving of story variants.

News-gathering infrastructure

Infrastructure is also required to take advantage of opportunities to use LLMs to substantially scale news-gathering by monitoring and analysing large volumes of source documents – perhaps as part of a differentiation strategy focused on competitively defensible coverage of a particular news domain. 

A starting point is infrastructure that enables constant access to source material, which might exist in different text formats, or in audio or video form that must first be converted to text transcripts using AI speech-to-text tools, or possibly even in large datasets that are frequently updated. Such ‘monitoring’ infrastructure might range from a simple list of sources to a complex web crawler that continually ‘reads’ portions of the internet to maintain an awareness of domain events. 

The monitoring that such infrastructure performs will depend on a system of prompts and a system that manages the summaries and assessments that are the outputs of such prompts. 

This all requires databases, data schemas, access tools and filtering functionality. Such functionality can be built at a small scale and used as a supplement to a manual monitoring workflow, and several newsrooms are already doing so.

Infrastructure flexibility

Despite our efforts to identify a useful way forward amid the uncertainty of the present moment, there remains a real possibility that AI might enable an entirely new interface to journalism, delivering experiences of news that are not centred on discrete text, audio or video artefacts. 

We see early signs of this in the form of chat interfaces, generative search and early conversational voice interfaces enabled by language models. Despite the caution provided above regarding an archive chatbot strategy, and despite evidence from audience research indicating a preference for passive rather than active news consumption, it is still possible that we may soon be interacting with news in entirely new ways. 

Given this uncertainty, it may be useful for news organisations to reexamine their technical architectures and infrastructure strategy from the perspective of an increasing need for flexibility. This might involve reexamining build-vs-buy decisions, refactoring brittle, tightly coupled architectures, or even attempting to identify various possible product scenarios and a path to infrastructure that could support those scenarios if they develop. 

Any such reevaluation should consider the many new options for AI-enabled software architectures and AI-enabled software engineering that are appearing. The increasing need for flexible infrastructure is accompanied by new AI-enabled techniques that may make it easier to design, build and maintain such infrastructure.

4. Organisational structure for AI-empowered teams

News organisations will need to operate differently

A news organisation with a coherent, well-articulated strategy for responding to generative AI, delivered via a portfolio of well-conceived applications supported by a professionalised prompt management process and delivered to audiences via a personalised publishing stack would clearly be well-placed to adapt to an AI-mediated information environment. 

But such an organisation would also likely need to operate differently from a traditional digital news publisher, and its organisational structure would probably need to change substantially in order to support those differences. Furthermore, the skills and talent required to operate successfully in this environment are likely to also be different. 

Discussions of likely changes in the structure of news organisations often focus on the potential for AI to either replace or augment traditional jobs. But the reality is likely to be more complicated than that. 

It is quite certain that many newsroom tasks will be either replaced or made moot by AI, but it is also certain that many new tasks will appear. It is obviously difficult at this stage to predict how old and new tasks will be assembled into individual jobs, teams and department-level functions, and different newsrooms with different strategies will clearly do this differently. Nonetheless we can make some informed speculation using the likely drivers of organisation change as a starting point.

Drivers of organisational change

If we consider what generative AI can do, how it is being used for news applications in a nascent way and how news organisations talk about it, then we can make some plausible assumptions about how news organisations might change in response to it:

• It is likely that technology will play a more central role in news organisations than it currently does. It is likely, perhaps less obviously, that the accessibility of generative AI will cause the use and control of technology to be dispersed throughout the organisation rather than concentrated within a team of specialists. 
• It is likely that news organisations will place a higher priority on adaptability and constant entrepreneurial innovation. 
• It is likely that substantial differences in the productivity of teams and individuals may appear, caused by differential effectiveness in using AI. 
• It is likely that, as competition for audience attention increases, news organisations will increasingly value a deep understanding of audiences and their information needs.
• It is likely that, at least for some categories of news, news organisations will focus less on producing individual stories by hand and more on overseeing systems and processes that produce or help produce stories. 

The collective organisational influence of all these assumptions can perhaps be summed up in a single word: autonomy. Those teams that adopt and master AI will be able to do much more, with many fewer dependencies on other parts of the organisation.

An AI-native news organisation

The enablement of increased autonomy by AI suggests to me that the productive units of AI-native news organisations might be small, AI-empowered, multi-disciplinary and self-directing teams operating relatively independently from each other and each focused on serving a specific audience or audience need. 

In this scenario the organisation itself becomes somewhat federalised, providing an environment within which self-directing teams can be productive and impactful but not directing their work. 

This federal organisation provides brand, values, certification of quality, monetisation, financial stability, enabling infrastructure, training and, of course, general strategy. The atomic teams provide adaptability, fast decision making, audience and competitive awareness and, of course, routine production of valued content via AI-augmented workflows. 

Implementing such a federalised organisational structure would clearly be challenging for many news organisations, not least because of that eternal and vaguely-defined bugbear of newsroom change management: culture. 

A practical, near-term response being explored by some pioneering newsrooms is the possibility of becoming a ‘two-speed organisation’, taking advantage of the new autonomy available from AI to set up teams that are loosely coupled to existing workflows. A longer-term response, already underway at several newsrooms, is to re-evaluate hiring and performance criteria to emphasise the skills and talents needed to form a more autonomous AI-empowered culture.

Skills and talent

The most valued skill in an AI-empowered news organisation will likely be the same as it has been in traditionally configured news organisations.

Editorial judgement – the ability to maintain a keen awareness of the deep informational needs of an audience or society, identify stories that meet those deep needs, verify and contextualise those stories, and then communicate them to audiences in clear and engaging forms – will probably remain the foundation of journalism. 

How editorial judgement is exercised, however, may change in ways that require substantial new skills and different talents. These might include an ability to work with abstractions and systems, to analytically understand audiences and their needs, to engage with complexity, and to remain curious and to learn continually. 

Organisations will obviously need to identify, support, incentivise and retain their most innovative and adaptable employees, but they will also need to supply those AI-empowered employees with leadership that is perhaps more entrepreneurial, more skilled at motivating and coaching, and perhaps less managerial or political. 

Specific technical skills with specific AI tools or models may, surprisingly, be less important due to the near-universal accessibility of the interfaces. Merely operating AI will likely be much easier and therefore less valuable than wielding it skilfully as a genuine superpower.

5. So, what’s next?

The AI moment

The development of generative AI has placed journalism at the cusp of significant change, variously equated to the iPhone moment, the birth of the internet and even the appearance of the printing press. 

The significance of the moment has been understood and appreciated by the senior leadership of most newsrooms, and many are already moving forward towards specific initiatives and experiments aimed at preparing for an AI-mediated future. 

Considerable attention is rightly being given to potential harms, to the ethics of using AI in journalism, to influencing regulation and legislation, to the potential for AI-created misinformation and disinformation, to education of audiences and ‘AI literacy’ and to the development of early guidelines to orient news organisations as they begin this new transformation of their industry. 

This attention is necessary and valuable, but it is not enough. Journalism must also engage with these new tools, explore them and their potential, and learn how to pragmatically apply them in creating and delivering value to audiences. 

There are no best practices, textbooks or shortcuts for this yet, only engaging, doing and learning until a viable way forward appears. Caution is advisable, but waiting for complete clarity is not. So-called ‘ second mover advantage ’ is only available to those who are well-prepared to move when the time comes.

The next information ecosystem

Looking further ahead, the need for hands-on familiarity with applied AI in journalism becomes even more critical because of the likelihood that the entire information ecosystem within which journalism exists will undergo transformation. 

What will journalism look like, for example, in an environment in which text, audio and video is fluid and malleable to the preferences of each individual consumer? What should the tangible output of a newsroom be in an environment in which that output is consumed primarily by machines? How will a coherent record of news – an archived ‘first draft of history’ – be maintained in such an environment? What might news become when useful reporting can be done on almost every word of text or speech, or every byte of data, produced in public by society? How will newsrooms capture value from their work in such an environment? What will that work be? 

These questions, and others that are similarly fundamental, may not become relevant for years, or even decades, but merely participating meaningfully in that discussion will require newsrooms to possess far more tangible expertise of AI-augmented journalism than any now possess. 

Progress in the face of uncertainty depends on developing and maintaining options, but options require situational awareness, and situational awareness comes from authentic engagement with the environment. For journalism that environment will almost certainly be shaped by AI.

In this article I have attempted to describe a few ways in which news organisations might build on the awareness of AI that they have developed since the launch of ChatGPT with specific strategies, projects, infrastructure and organisational changes. 

These suggestions are my own interpretation of what I am observing and hearing, and they are by no means exhaustive or complete. It would be reasonable to expect people who staff or lead news organisations to exhibit frustration or even resentment at the prospect of even more impending change, but this is not what I have observed. 

Instead the predominant tone of my conversations about AI has been one of optimism and excitement, largely at the opportunity presented by AI to further realise an ideal of journalism that motivates many of us who work in and around newsrooms. If you could bring that ideal to life, without regard for the scarcity of resources, where would you start?

David Caswell is a consultant, builder and researcher focused on AI in news. David has led news product innovation at the BBC, Tribune Publishing & Yahoo! and publish peer-reviewed work. He is a regular speaker of our newsroom leadership programmes . This piece, which was written entirely by hand and without the use of AI, was first published here .

At every email we send you'll find original reporting, evidence-based insights, online seminars and readings curated from 100s sources - all in 5 minutes.

• Twice a week
• More than 20,000 people receive it
• Unsubscribe any time

signup block

How the news ecosystem might look like in the age of generative ai.

Is ChatGPT a threat or an opportunity for journalism? Five AI experts weigh in

Will AI-generated images create a new crisis for fact-checkers? Experts are not so sure

AI, lies and conspiracy theories: How Latinos became a key target for misinformation in the US election

AI deepfakes, bad laws – and a big fat Indian election

Latest News

Find anything you save across the site in your account

Does Journalism Have a Future?

By Jill Lepore

The wood-panelled tailgate of the 1972 Oldsmobile station wagon dangled open like a broken jaw, making a wobbly bench on which four kids could sit, eight legs swinging. Every Sunday morning, long before dawn, we’d get yanked out of bed to stuff the car’s way-back with stacks of twine-tied newspapers, clamber onto the tailgate, cut the twine with my mother’s sewing scissors, and ride around town, bouncing along on that bench, while my father shouted out orders from the driver’s seat. “Watch out for the dog!” he’d holler between draws on his pipe. “Inside the screen door!” “Mailbox!” As the car crept along, never stopping, we’d each grab a paper and dash in the dark across icy driveways or dew-drunk grass, crashing, seasonally, into unexpected snowmen. “Back porch!” “Money under the mat!” He kept a list, scrawled on the back of an envelope, taped to the dashboard: the Accounts. “They owe three weeks!” He didn’t need to remind us. We knew each Doberman and every debt. We’d deliver our papers—Worcester Sunday Telegrams —and then run back to the car and scramble onto the tailgate, dropping the coins we’d collected into empty Briggs tobacco tins as we bumped along to the next turn, the newspaper route our Sabbath.

The Worcester Sunday Telegram was founded in 1884, when a telegram meant something fast. Two years later, it became a daily. It was never a great paper but it was always a pretty good paper: useful, gossipy, and resolute. It cultivated talent. The poet Stanley Kunitz was a staff writer for the Telegram in the nineteen-twenties. The New York Times reporter Douglas Kneeland, who covered Kent State and Charles Manson, began his career there in the nineteen-fifties. Joe McGinniss reported for the Telegram in the nineteen-sixties before writing “The Selling of the President.” From bushy-bearded nineteenth-century politicians to baby-faced George W. Bush, the paper was steadfastly Republican, if mainly concerned with scandals and mustachioed villains close to home: overdue repairs to the main branch of the public library, police raids on illegal betting establishments—“ Worcester Dog Chases Worcester Cat Over Worcester Fence ,” as the old Washington press-corps joke about a typical headline in a local paper goes. Its pages rolled off giant, thrumming presses in a four-story building that overlooked City Hall the way every city paper used to look out over every city hall, the Bat-Signal over Gotham.

Most newspapers like that haven’t lasted. Between 1970 and 2016, the year the American Society of News Editors quit counting, five hundred or so dailies went out of business; the rest cut news coverage, or shrank the paper’s size, or stopped producing a print edition, or did all of that, and it still wasn’t enough. The newspaper mortality rate is old news, and nostalgia for dead papers is itself pitiful at this point, even though, I still say, there’s a principle involved. “I wouldn’t weep about a shoe factory or a branch-line railroad shutting down,” Heywood Broun, the founder of the American Newspaper Guild, said when the New York World went out of business, in 1931. “But newspapers are different.” And the bleeding hasn’t stopped. Between January, 2017, and April, 2018, a third of the nation’s largest newspapers, including the Denver Post and the San Jose Mercury News , reported layoffs. In a newer trend, so did about a quarter of digital-native news sites. BuzzFeed News laid off a hundred people in 2017; speculation is that BuzzFeed is trying to dump it. The Huffington Post paid most of its writers nothing for years, upping that recently to just above nothing, and yet, despite taking in tens of millions of dollars in advertising revenue in 2018, it failed to turn a profit.

Even veterans of august and still thriving papers are worried, especially about the fake news that’s risen from the ashes of the dead news. “We are, for the first time in modern history, facing the prospect of how societies would exist without reliable news,” Alan Rusbridger, for twenty years the editor-in-chief of the Guardian , writes in “ Breaking News: The Remaking of Journalism and Why It Matters Now .” “There are not that many places left that do quality news well or even aim to do it at all,” Jill Abramson , a former executive editor of the New York Times , writes in “Merchants of Truth: The Business of News and the Fight for Facts.” Like most big-paper reporters and editors who write about the crisis of journalism, Rusbridger and Abramson are interested in national and international news organizations. The local story is worse.

First came conglomeration. Worcester, Massachusetts, the second-largest city in New England, used to have four dailies: the Telegram , in the morning, and the Gazette , in the evening (under the same ownership), the Spy , and the Post . Now it has one. The last great laying waste to American newspapers came in the early decades of the twentieth century, mainly owing to (a) radio and (b) the Depression; the number of dailies fell from 2,042 in 1920 to 1,754 in 1944, leaving 1,103 cities with only one paper. Newspaper circulation rose between 1940 and 1990, but likely only because more people were reading fewer papers, and, as A. J. Liebling once observed, nothing is crummier than a one-paper town. In 1949, after yet another New York daily closed its doors, Liebling predicted, “If the trend continues, New York will be a one- or two-paper town by about 1975.” He wasn’t that far off. In the nineteen-eighties and nineties, as Christopher B. Daly reports in “ Covering America: A Narrative History of the Nation’s Journalism ,” “the big kept getting bigger.” Conglomeration can be good for business, but it has generally been bad for journalism. Media companies that want to get bigger tend to swallow up other media companies, suppressing competition and taking on debt, which makes publishers cowards. In 1986, the publisher of the San Francisco Chronicle bought the Worcester Telegram and the Evening Gazette , and, three years later, right about when Time and Warner became Time Warner, the Telegram and the Gazette became the Telegram & Gazette , or the T&G , smaller fries but the same potato.

Next came the dot-coms. Craigslist went online in the Bay Area in 1996 and spread across the continent like a weed, choking off local newspapers’ most reliable source of revenue: classified ads. The T&G tried to hold on to its classified-advertising section by wading into the shallow waters of the Internet, at telegram.com, where it was called, acronymically, and not a little desperately, “ TANGO !” Then began yet another round of corporate buyouts, deeply leveraged deals conducted by executives answerable to stockholders seeking higher dividends, not better papers. In 1999, the New York Times Company bought the T&G for nearly three hundred million dollars. By 2000, only three hundred and fifty of the fifteen hundred daily newspapers left in the United States were independently owned. And only one out of every hundred American cities that had a daily newspaper was anything other than a one-paper town.

Then came the fall, when papers all over the country, shackled to mammoth corporations and a lumbering, century-old business model, found themselves unable to compete with the upstarts—online news aggregators like the Huffington Post (est. 2005) and Breitbart News (est. 2007), which were, to readers, free. News aggregators also drew display advertisers away from print; Facebook and Google swallowed advertising accounts whole. Big papers found ways to adapt; smaller papers mainly folded. Between 1994 and 2016, years when the population of Worcester County rose by more than a hundred thousand, daily home delivery of the T&G declined from more than a hundred and twenty thousand to barely thirty thousand. In one year alone, circulation fell by twenty-nine per cent. In 2012, after another round of layoffs, the T&G left its building, its much reduced staff small enough to fit into two floors of an office building nearby. The next year, the owner of the Boston Red Sox bought the newspaper, along with the Boston Globe , from the New York Times Company for seventy million dollars, only to unload the T&G less than a year later, for seventeen million dollars, to Halifax Media Group, which held it for only half a year before Halifax itself was bought, flea-market style, by an entity that calls itself, unironically, the New Media Investment Group.

The numbers mask an uglier story. In the past half century, and especially in the past two decades, journalism itself—the way news is covered, reported, written, and edited—has changed, including in ways that have made possible the rise of fake news, and not only because of mergers and acquisitions, and corporate ownership, and job losses, and Google Search, and Facebook and BuzzFeed . There’s no shortage of amazing journalists at work, clear-eyed and courageous, broad-minded and brilliant, and no end of fascinating innovation in matters of form, especially in visual storytelling. Still, journalism, as a field, is as addled as an addict, gaunt, wasted, and twitchy, its pockets as empty as its nights are sleepless. It’s faster than it used to be, so fast. It’s also edgier, and needier, and angrier. It wants and it wants and it wants. But what does it need?

The daily newspaper is the taproot of modern journalism. Dailies mainly date to the eighteen-thirties, the decade in which the word “journalism” was coined, meaning daily reporting, the jour in journalism. Early dailies depended on subscribers to pay the bills. The press was partisan, readers were voters, and the news was meant to persuade (and voter turnout was high). But by 1900 advertising made up more than two-thirds of the revenue at most of the nation’s eighteen thousand newspapers, and readers were consumers (and voter turnout began its long fall). “The newspaper is not a missionary or a charitable institution, but a business that collects and publishes news which the people want and are willing to buy,” one Missouri editor said in 1892. Newspapers stopped rousing the rabble so much because businesses wanted readers, no matter their politics. “There is a sentiment gaining ground to the effect that the public wants its politics ‘straight,’ ” a journalist wrote the following year. Reporters pledged themselves to “facts, facts, and more facts,” and, as the press got less partisan and more ad-based, newspapers sorted themselves out not by their readers’ political leanings but by their incomes. If you had a lot of money to spend, you read the St. Paul Pioneer Press; if you didn’t have very much, you read the St. Paul Dispatch .

Unsurprisingly, critics soon began writing big books, usually indictments, about the relationship between business and journalism. “When you read your daily paper, are you reading facts or propaganda?” Upton Sinclair asked on the jacket of “ The Brass Check ,” in 1919. In “ The Disappearing Daily ,” in 1944, Oswald Garrison Villard mourned “what was once a profession but is now a business.” The big book that inspired Jill Abramson to become a journalist was David Halberstam’s “ The Powers That Be ,” from 1979, a history of the rise of the modern, corporate-based media in the middle decades of the twentieth century. Halberstam, who won a Pulitzer Prize in 1964 for his reporting from Vietnam for the New York Times , took up his story more or less where Villard left off. He began with F.D.R. and CBS radio; added the Los Angeles Times , Time Inc., and CBS television; and reached his story’s climax with the Washington Post and the New York Times and the publication of the Pentagon Papers, in 1971.

Halberstam argued that between the nineteen-thirties and the nineteen-seventies radio and television brought a new immediacy to reporting, while the resources provided by corporate owners and the demands made by an increasingly sophisticated national audience led to harder-hitting, investigative, adversarial reporting, the kind that could end a war and bring down a President. Richard Rovere summed it up best: “What The Los Angeles Times, The Washington Post, Time and CBS have in common is that, under pressures generated internally and externally, they moved from venality or parochialism or mediocrity or all three to something approaching journalistic excellence and responsibility.” That move came at a price. “Watergate, like Vietnam, had obscured one of the central new facts about the role of journalism in America,” Halberstam wrote. “Only very rich, very powerful corporate institutions like these had the impact, the reach, and above all the resources to challenge the President of the United States.”

Link copied

There’s reach, and then there’s reach. When I was growing up, in the nineteen-seventies, nobody I knew read the New York Times , the Washington Post , or the Wall Street Journal . Nobody I knew even read the Boston Globe , a paper that used to have a rule that no piece should ever be so critical of anyone that its “writer could not shake hands the next day with the man about whom he had written.” After journalism put up its dukes, my father only ever referred to the Globe as “that Communist rag,” not least because, in 1967, it became the first major paper in the United States to come out against the Vietnam War.

The view of the new journalism held by people like my father escaped Halberstam’s notice. In 1969, Nixon’s Vice-President, Spiro Agnew, delivered a speech drafted by the Nixon aide Pat Buchanan accusing the press of liberal bias. It’s “good politics for us to kick the press around,” Nixon is said to have told his staff. The press, Agnew said, represents “a concentration of power over American public opinion unknown in history,” consisting of men who “read the same newspapers” and “talk constantly to one another.” How dare they. Halberstam waved this aside as so much P.R. hooey, but, as has since become clear, Agnew reached a ready audience, especially in houses like mine.

Spiro who? “The press regarded Agnew with uncontrolled hilarity,” Arthur Schlesinger, Jr., observed in 1970, but “no one can question the force of Spiro T. Agnew’s personality, nor the impact of his speeches.” No scholar of journalism can afford to ignore Agnew anymore. In “ On Press: The Liberal Values That Shaped the News ,” the historian Matthew Pressman argues that any understanding of the crisis of journalism in the twenty-first century has to begin by vanquishing the ghost of Spiro T. Agnew.

For Pressman, the pivotal period for the modern newsroom is what Abramson calls “Halberstam’s Golden Age,” between 1960 and 1980, and its signal feature was the adoption not of a liberal bias but of liberal values: “Interpretation replaced transmission, and adversarialism replaced deference.” In 1960, nine out of every ten articles in the Times about the Presidential election were descriptive; by 1976, more than half were interpretative. This turn was partly a consequence of television—people who simply wanted to find out what happened could watch television, so newspapers had to offer something else—and partly a consequence of McCarthyism . “The rise of McCarthy has compelled newspapers of integrity to develop a form of reporting which puts into context what men like McCarthy have to say,” the radio commentator Elmer Davis said in 1953. Five years later, the Times added “News Analysis” as a story category. “Once upon a time, news stories were like tape recorders,” the Bulletin of the American Society of Newspaper Editors commented in 1963. “No more. A whole generation of events had taught us better—Hitler and Goebbels, Stalin and McCarthy, automation and analog computers and missiles.”

These changes weren’t ideologically driven, Pressman insists, but they had ideological consequences. At the start, leading conservatives approved. “To keep a reporter’s prejudices out of a story is commendable,” Irving Kristol wrote in 1967. “To keep his judgment out of a story is to guarantee that the truth will be emasculated.” After the Times and the Post published the Pentagon Papers, Kristol changed his spots. Journalists, he complained in 1972, were now “engaged in a perpetual confrontation with the social and political order (the ‘establishment,’ as they say).” By 1975, after Watergate, Kristol was insisting that “most journalists today . . . are ‘liberals.’ ” With that, the conservative attack on the press was off and running, all the way to Trumpism—“the failing New York Times,” “CNN is fake news,” the press is “the true enemy of the people”—and, in a revolution-devouring-its-elders sort of way, the shutting down of William Kristol’s Weekly Standard , in December. “The pathetic and dishonest Weekly Standard . . . is flat broke and out of business,” Trump tweeted. “May it rest in peace!”

What McCarthy and television were for journalism in the nineteen-fifties, Trump and social media would be in the twenty-tens: license to change the rules. Halberstam’s Golden Age, or what he called “journalism’s high-water mark,” ended about 1980. Abramson’s analysis in “Merchants of Truth” begins with journalism’s low-water mark, in 2007, the year after Facebook launched its News Feed, “the year everything began to fall apart.”

“ Merchants of Truth ” isn’t just inspired by “The Powers That Be”; it’s modelled on it. Abramson’s book follows Halberstam’s structure and mimics its style, chronicling the history of a handful of nationally prominent media organizations—in her case, BuzzFeed, Vice, the Times , and the Washington Post —in alternating chapters that are driven by character sketches and reported scenes. The book is saturated with a lot of gossip and glitz, including details about the restaurants the powers that be frequent, and what they wear (“Sulzberger”—the Times ’ publisher—“dressed in suits from Bloomingdale’s, stylish without being ostentatiously bespoke, and wore suspenders before they went out of fashion”), alongside crucial insights about structural transformations, like how Web and social-media publishing “unbundled” the newspaper, so that readers who used to find a fat newspaper on their front porch could, on their phones, look, instead, at only one story. “Each individual article now lived on its own page, where it had a unique URL and could be shared, and spread virally,” Abramson observes. “This put stories, rather than papers, in competition with one another.”

This history is a chronicle of missed opportunities, missteps, and lessons learned the hard way. As long ago as 1992, an internal report at the Washington Post urged the mounting of an “electronic product”: “The Post ought to be in the forefront of this.” Early on, the Guardian started a New Media lab, which struck a lot of people as frivolous, Rusbridger writes, because, at the time, “only 3 per cent of households owned a PC and a modem,” a situation not unlike that at the Guardian’s own offices, where “it was rumored that downstairs a bloke called Paul in IT had a Mac connected to the internet.” A 1996 business plan for the Guardian concluded that the priority was print, and the London Times editor Simon Jenkins predicted, “The Internet will strut an hour upon the stage, and then take its place in the ranks of the lesser media.” In 2005, the Post lost a chance at a ten-per-cent investment in Facebook, whose returns, as Abramson points out, would have floated the newspaper for decades. The C.E.O. of the Washington Post Company, Don Graham, and Mark Zuckerberg shook hands over the deal, making a verbal contract, but, when Zuckerberg weaseled out of it to take a better offer, Graham, out of kindness to a young fella just starting out, simply let him walk away. The next year, the Post shrugged off a proposal from two of its star political reporters to start a spinoff Web site; they went on to found Politico. The Times , Abramson writes, declined an early chance to invest in Google, and was left to throw the kitchen sink at its failing business model, including adding a Thursday Style section to attract more high-end advertising revenue. Bill Keller, then the newspaper’s editor, said, “If luxury porn is what saves the Baghdad bureau, so be it.”

More alarming than what the Times and the Post failed to do was how so much of what they did do was determined less by their own editors than by executives at Facebook and BuzzFeed. If journalism has been reinvented during the past two decades, it has, in the main, been reinvented not by reporters and editors but by tech companies, in a sequence of events that, in Abramson’s harrowing telling, resemble a series of puerile stunts more than acts of public service.

Who even are these people? “Merchants of Truth” has been charged with factual errors, including by people Abramson interviewed, especially younger journalists. She can also be maddeningly condescending. She doffs her cap at Sulzberger, with his natty suspenders, but dismisses younger reporters at places like Vice as notable mainly for being “impossibly hip, with interesting hair.” This is distracting, and too bad, because there is a changing of the guard worth noting, and it’s not incidental: it’s critical. All the way through to the nineteen-eighties, all sorts of journalists, including magazine, radio, and television reporters, got their start working on daily papers, learning the ropes and the rules. Rusbridger started out in 1976 as a reporter at the Cambridge Evening News , which covered stories that included a petition about a pedestrian crossing and a root vegetable that looked like Winston Churchill. In the U.K., a reporter who wanted to go to Fleet Street had first to work for three years on a provincial newspaper, pounding the pavement. Much the same applied in the U.S., where a cub reporter did time at the Des Moines Register , or the Worcester Telegram , before moving up to the New York Times or the Herald Tribune. Beat reporting, however, is not the backstory of the people who, beginning in the nineteen-nineties, built the New Media.

Jonah Peretti started out soaking up postmodern theory at U.C. Santa Cruz in the mid-nineteen-nineties, and later published a scholarly journal article about the scrambled, disjointed, and incoherent way of thinking produced by accelerated visual experiences under late capitalism. Or something like that. Imagine an article written by that American Studies professor in Don DeLillo’s “ White Noise .” Peretti thought that watching a lot of MTV can mess with your head—“The rapid fire succession of signifiers in MTV style media erodes the viewer’s sense of temporal continuity”—leaving you confused, stupid, and lonely. “Capitalism needs schizophrenia, but it also needs egos,” Peretti wrote. “The contradiction is resolved through the acceleration of the temporal rhythm of late capitalist visual culture. This type of acceleration encourages weak egos that are easily formed, and fade away just as easily.” Voilà, a business plan!

Peretti’s career in viral content began in 2001, with a prank involving e-mail and Nike sneakers while he was a graduate student at the M.I.T. Media Lab. (Peretti ordered custom sneakers embroidered with the word “sweatshop” and then circulated Nike’s reply.) In 2005, a year the New York Times Company laid off five hundred employees and the Post began paying people to retire early, Peretti joined Andrew Breitbart, a Matt Drudge acolyte, and Ken Lerer, a former P.R. guy at AOL Time Warner, in helping Arianna Huffington, a millionaire and a former anti-feminist polemicist, launch the Huffington Post. Peretti was in charge of innovations that included a click-o-meter. Within a couple of years, the Huffington Post had more Web traffic than the Los Angeles Times , the Washington Post , and the Wall Street Journal . Its business was banditry. Abramson writes that when the Times published a deeply reported exclusive story about WikiLeaks, which took months of investigative work and a great deal of money, the Huffington Post published its own version of the story, using the same headline—and beat out the Times story in Google rankings. “We were learning that the internet behaved like a clattering of jackdaws,” Rusbridger writes. “Nothing remained exclusive for more than two minutes.”

Pretty soon, there were jackdaws all over the place, with their schizophrenic late-capitalist accelerated signifiers. Breitbart left the Huffington Post and started Breitbart News around the same time that Peretti left to focus on his own company, Contagious Media, from which he launched BuzzFeed, where he tested the limits of virality with offerings like the seven best links about gay penguins and “YouTube Porn Hacks.” He explained his methods in a pitch to venture capitalists: “Raw buzz is automatically published the moment it is detected by our algorithm,” and “the future of the industry is advertising as content.”

Facebook launched its News Feed in 2006. In 2008, Peretti mused on Facebook, “Thinking about the economics of the news business.” The company added its Like button in 2009. Peretti set likability as BuzzFeed’s goal, and, to perfect the instruments for measuring it, he enlisted partners, including the Times and the Guardian , to share their data with him in exchange for his reports on their metrics. Lists were liked. Hating people was liked. And it turned out that news, which is full of people who hate other people, can be crammed into lists.

Chartbeat, a “content intelligence” company founded in 2009, launched a feature called Newsbeat in 2011. Chartbeat offers real-time Web analytics, displaying a constantly updated report on Web traffic that tells editors what stories people are reading and what stories they’re skipping. The Post winnowed out reporters based on their Chartbeat numbers. At the offices of Gawker, the Chartbeat dashboard was displayed on a giant screen.

In 2011, Peretti launched BuzzFeed News, hiring a thirty-five-year-old Politico journalist, Ben Smith, as its editor-in-chief. Smith asked for a “scoop-a-day” from his reporters, who, he told Abramson, had little interest in the rules of journalism: “They didn’t even know what rules they were breaking.” In 2012, BuzzFeed introduced three new one-click ways for readers to respond to stories, beyond “liking” them—LOL, OMG, and WTF—and ran lists like “10 Reasons Everyone Should Be Furious About Trayvon Martin’s Murder,” in which, as Abramson explains, BuzzFeed “simply lifted what it needed from reports published elsewhere, repackaged the information, and presented it in a way that emphasized sentiment and celebrity.” BuzzFeed makes a distinction between BuzzFeed and BuzzFeed News, just as newspapers and magazines draw distinctions between their print and their digital editions. These distinctions are lost on most readers. BuzzFeed News covered the Trayvon Martin story, but its information, like BuzzFeed’s, came from Reuters and the Associated Press.

Even as news organizations were pruning reporters and editors, Facebook was pruning its users’ news, with the commercially appealing but ethically indefensible idea that people should see only the news they want to see. In 2013, Silicon Valley began reading its own online newspaper, the Information, its high-priced subscription peddled to the information élite, following the motto “Quality stories breed quality subscribers.” Facebook’s goal, Zuckerberg explained in 2014, was to “build the perfect personalized newspaper for every person in the world.” Ripples at Facebook create tsunamis in newsrooms. The ambitious news site Mic relied on Facebook to reach an audience through a video program called Mic Dispatch, on Facebook Watch; last fall, after Facebook suggested that it would drop the program, Mic collapsed. Every time Facebook News tweaks its algorithm—tweaks made for commercial, not editorial, reasons—news organizations drown in the undertow. An automated Facebook feature called Trending Topics, introduced in 2014, turned out to mainly identify junk as trends, and so “news curators,” who tended to be recent college graduates, were given a new, manual mandate, “massage the algorithm,” which meant deciding, themselves, which stories mattered. The fake news that roiled the 2016 election? A lot of that was stuff on Trending Topics. (Last year, Facebook discontinued the feature.)

BuzzFeed surpassed the Times Web site in reader traffic in 2013. BuzzFeed News is subsidized by BuzzFeed, which, like many Web sites—including, at this point, those of most major news organizations—makes money by way of “native advertising,” ads that look like articles. In some publications, these fake stories are easy to spot; in others, they’re not. At BuzzFeed, they’re in the same font as every other story. BuzzFeed’s native-advertising bounty meant that BuzzFeed News had money to pay reporters and editors, and it began producing some very good and very serious reporting, real news having become something of a luxury good. By 2014, BuzzFeed employed a hundred and fifty journalists, including many foreign correspondents. It was obsessed with Donald Trump’s rumored Presidential bid, and followed him on what it called the “fake campaign trail” as early as January, 2014. “It used to be the New York Times , now it’s BuzzFeed,” Trump said, wistfully. “The world has changed.” At the time, Steve Bannon was stumping for Trump on Breitbart. Left or right, a Trump Presidency was just the sort of story that could rack up the LOLs, OMGs, and WTFs. It still is.

In March, 2014, the Times produced an Innovation Report, announcing that the newspaper had fallen behind in “the art and science of getting our journalism to readers,” a field led by BuzzFeed. That May, Sulzberger fired Abramson, who had been less than all-in about the Times doing things like running native ads. Meanwhile, BuzzFeed purged from its Web site more than four thousand of its early stories. “It’s stuff made at a time when people were really not thinking of themselves as doing journalism,” Ben Smith explained. Not long afterward, the Times began running more lists, from book recommendations to fitness tips to takeaways from Presidential debates.

The Times remains unrivalled. It staffs bureaus all over the globe and sends reporters to some of the world’s most dangerous places. It has more than a dozen reporters in China alone. Nevertheless, BuzzFeed News became more like the Times , and the Times became more like BuzzFeed, because readers, as Chartbeat announced on its endlessly flickering dashboards, wanted lists, and luxury porn, and people to hate.

The Guardian , founded as the Manchester Guardian in 1821, has been held by a philanthropic trust since 1936, which somewhat insulates it from market forces, just as Jeff Bezos’s ownership now does something similar for the Post . By investing in digital-readership research from the time Rusbridger took charge, in 1995, the Guardian became, for a while, the online market leader in the U.K. By 2006, two-thirds of its digital readers were outside the U.K. In 2007, the Guardian undertook what Rusbridger calls “the Great Integration,” pulling its Web and print parts together into a single news organization, with the same editorial management. It also developed a theory about the relationship between print and digital, deciding, in 2011, to be a “digital-first organization” and to “make print a slower, more reflective read which would not aspire to cover the entire waterfront in news.”

Rusbridger explains, with a palpable grief, his dawning realization that the rise of social media meant that “chaotic information was free: good information was expensive,” which meant, in turn, that “good information was increasingly for smaller elites” and that “it was harder for good information to compete on equal terms with bad.” He takes these circumstances as something of a dare: “Our generation had been handed the challenge of rethinking almost everything societies had, for centuries, taken for granted about journalism.”

Has that challenge been met? The Guardian’s own success is mixed. As of 2018, it was in the black, partly by relying on philanthropy, especially in the U.S. “Reader revenue,” in the form of donations marked not as subscriptions but as voluntary “memberships,” is expected to overtake advertising revenue before long. Raising money from people who care about journalism has allowed the Guardian to keep the Web site free. It’s also broken some big stories, from the Murdoch-papers phone-hacking scoop to the saga of Edward Snowden, and provided riveting coverage of ongoing and urgent stories, especially climate change. But, for all its fine reporting and substantive “Long Reads,” the paper consists disproportionately of ideologically unvarying opinion essays. By some measures, journalism entered a new, Trumpian, gold-plated age during the 2016 campaign, with the Trump bump, when news organizations found that the more they featured Trump the better their Chartbeat numbers, which, arguably, is a lot of what got him elected. The bump swelled into a lump and, later, a malignant tumor, a carcinoma the size of Cleveland. Within three weeks of the election, the Times added a hundred and thirty-two thousand new subscribers. (This effect hasn’t extended to local papers.) News organizations all over the world now advertise their services as the remedy to Trumpism, and to fake news; fighting Voldemort and his Dark Arts is a good way to rake in readers. And scrutiny of the Administration has produced excellent work, the very best of journalism. “ How President Trump Is Saving Journalism ,” a 2017 post on Forbes.com, marked Trump as the Nixon to today’s rising generation of Woodwards and Bernsteins. Superb investigative reporting is published every day, by news organizations both old and new, including BuzzFeed News.

By the what-doesn’t-kill-you line of argument, the more forcefully Trump attacks the press, the stronger the press becomes. Unfortunately, that’s not the full story. All kinds of editorial decisions are now outsourced to Facebook’s News Feed, Chartbeat, or other forms of editorial automation, while the hands of many flesh-and-blood editors are tied to so many algorithms. For one reason and another, including twenty-first-century journalism’s breakneck pace, stories now routinely appear that might not have been published a generation ago, prompting contention within the reportorial ranks. In 2016, when BuzzFeed News released the Steele dossier, many journalists disapproved, including CNN’s Jake Tapper, who got his start as a reporter for the Washington City Paper . “It is irresponsible to put uncorroborated information on the Internet,” Tapper said. “It’s why we did not publish it, and why we did not detail any specifics from it, because it was uncorroborated, and that’s not what we do.” The Times veered from its normal practices when it published an anonymous opinion essay by a senior official in the Trump Administration. And The New Yorker posted a story online about Brett Kavanaugh’s behavior when he was an undergraduate at Yale, which Republicans in the Senate pointed to as evidence of a liberal conspiracy against the nominee.

There’s plenty of room to argue over these matters of editorial judgment. Reasonable people disagree. Occasionally, those disagreements fall along a generational divide. Younger journalists often chafe against editorial restraint, not least because their cohort is far more likely than senior newsroom staff to include people from groups that have been explicitly and viciously targeted by Trump and the policies of his Administration, a long and growing list that includes people of color, women, immigrants, Muslims, members of the L.G.B.T.Q. community, and anyone with family in Haiti or any of the other countries Trump deems “shitholes.” Sometimes younger people are courageous and sometimes they are heedless and sometimes those two things are the same. “The more ‘woke’ staff thought that urgent times called for urgent measures,” Abramson writes, and that “the dangers of Trump’s presidency obviated the old standards.” Still, by no means is the divide always or even usually generational. Abramson, for instance, sided with BuzzFeed News about the Steele dossier, just as she approves of the use of the word “lie” to refer to Trump’s lies, which, by the Post’s reckoning, came at the rate of more than a dozen a day in 2018.

The broader problem is that the depravity, mendacity, vulgarity, and menace of the Trump Administration have put a lot of people, including reporters and editors, off their stride. The present crisis, which is nothing less than a derangement of American life, has caused many people in journalism to make decisions they regret, or might yet. In the age of Facebook, Chartbeat, and Trump, legacy news organizations, hardly less than startups, have violated or changed their editorial standards in ways that have contributed to political chaos and epistemological mayhem. Do editors sit in a room on Monday morning, twirl the globe, and decide what stories are most important? Or do they watch Trump’s Twitter feed and let him decide? It often feels like the latter. Sometimes what doesn’t kill you doesn’t make you stronger; it makes everyone sick. The more adversarial the press, the more loyal Trump’s followers, the more broken American public life. The more desperately the press chases readers, the more our press resembles our politics.

The problems are well understood, the solutions harder to see. Good reporting is expensive, but readers don’t want to pay for it. The donation-funded ProPublica, “an independent, nonprofit newsroom that produces investigative journalism with moral force,” employs more than seventy-five journalists. Good reporting is slow, good stories unfold, and most stories that need telling don’t involve the White House. The Correspondent, an English-language version of the Dutch Web site De Correspondent, is trying to “unbreak the news.” It won’t run ads. It won’t collect data (or, at least, not much). It won’t have subscribers. Like NPR, it will be free for everyone, supported by members, who pay what they can. “We want to radically change what news is about, how it is made, and how it is funded,” its founders state. Push-notifications-on news is bad for you, they say, “because it pays more attention to the sensational, exceptional, negative, recent, and incidental, thereby losing sight of the ordinary, usual, positive, historical, and systematic.” What will the Correspondent look like? It will stay above the fray. It might sometimes be funny. It’s slated to début sometime in 2019. Aside from the thing about ads, it sounds a lot like a magazine, when magazines came in the mail.

After we’d shoved the last, fat Worcester Sunday Telegram inside the last, unlatched screen door, we’d head home, my father taking turns a little too fast, so that we’d have to clutch at one another and at the lip of the tailgate, to keep from falling off. “Dad, slow down!” we’d squeal, not meaning it. Then he’d make breakfast, hot chocolate with marshmallows in the winter, orange juice from a can of frozen concentrate in the summer, and on my plate I’d make wedges of cantaloupe into Viking ships sailing across a sea of maple syrup from the Coast of Bacon to Pancake Island. After breakfast, we’d dump the money from the tobacco tins onto the kitchen table and count coins, stacking quarters and nickels and dimes into wrappers from the Worcester County Institution for Savings, while my father updated the Accounts, and made the Collection List.

Going collecting was a drag. You had to knock on people’s doors and ask your neighbors for money—“ Telegram! Collecting!”—and it was embarrassing, and, half the time, they’d ask you in, and before you knew it you’d be helping out, and it would take all day. “So long as you’re here, could you hold the baby while I take a quick shower?” “Honey, after this, could you bring my mail down to the post office on that cute little bike of yours?” I came to understand that the people who didn’t leave the money under the mat hadn’t forgotten to. They just liked having a kid visit on Sunday afternoon.

The death of a newspaper is sometimes like other deaths. The Mrs. and the Miss, a very, very old woman and her very old daughter, lived in a crooked green house on top of a rise and wore matching housecoats and slippers. The Miss followed the Mrs. around like a puppy, and, if you found them in the parlor reading the paper, the Mrs. would be poring over the opinion pages while the Miss cut pictures out of the funnies. “The Miss can’t think straight,” my father said. “Her head’s scrambled. So be gentle with her. Nothing to be afraid of. Be sure to help them out.” Once when I biked over there, the Miss was standing, keening, noise without words, sound without sense. The Mrs. wasn’t moving, and she wasn’t ever going to move again. I called for help and held the Miss’s hand, waiting for the wail of sirens. I didn’t know what else to do. ♦

An earlier version of this story misstated the subtitle of Christopher B. Daly’s book “Covering America.” It also misstated the Huffington Post’s advertising revenue.

By signing up, you agree to our User Agreement and Privacy Policy & Cookie Statement . This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

By Ken Auletta

By Nicholas Lemann

Five things everybody needs to know about the future of Journalism

What do we know so far about the implications for journalism and by extension politics? Image:  REUTERS/Jonathan Ernst

.chakra .wef-1c7l3mo{-webkit-transition:all 0.15s ease-out;transition:all 0.15s ease-out;cursor:pointer;-webkit-text-decoration:none;text-decoration:none;outline:none;color:inherit;}.chakra .wef-1c7l3mo:hover,.chakra .wef-1c7l3mo[data-hover]{-webkit-text-decoration:underline;text-decoration:underline;}.chakra .wef-1c7l3mo:focus,.chakra .wef-1c7l3mo[data-focus]{box-shadow:0 0 0 3px rgba(168,203,251,0.5);} Rasmus Kleis Nielsen

Meera selva.

.chakra .wef-9dduvl{margin-top:16px;margin-bottom:16px;line-height:1.388;font-size:1.25rem;}@media screen and (min-width:56.5rem){.chakra .wef-9dduvl{font-size:1.125rem;}} Explore and monitor how .chakra .wef-15eoq1r{margin-top:16px;margin-bottom:16px;line-height:1.388;font-size:1.25rem;color:#F7DB5E;}@media screen and (min-width:56.5rem){.chakra .wef-15eoq1r{font-size:1.125rem;}} Digital Communications is affecting economies, industries and global issues

.chakra .wef-1nk5u5d{margin-top:16px;margin-bottom:16px;line-height:1.388;color:#2846F8;font-size:1.25rem;}@media screen and (min-width:56.5rem){.chakra .wef-1nk5u5d{font-size:1.125rem;}} Get involved with our crowdsourced digital platform to deliver impact at scale

Stay up to date:, digital communications.

What do we know about the future of journalism?

The rise of digital media has empowered people worldwide but also enabled the spread of disinformation and demagoguery and undermined the funding of professional journalism as we know it.

The move from a media environment defined by broadcasting and newspapers to a digital, mobile, and platform-dominated environment is the most fundamental change in how we communicate since the development of the printing press, and we are only thirty years into a period we can trace back to Tim Berners-Lee’s invention of the World Wide Web in 1989.

What do we know so far about the implications for journalism and by extension politics?

Some key broad trends can be clearly documented, and they are sometimes at odds with much of what is asserted in public and elite debate.

Too often, discussions of the future of media are based on misunderstandings or outright “media change denial” where people double down on arguments that are directly contradicted by a growing consensus among researchers.

Here are five things we believe everybody needs to know about the future of journalism , all backed by evidence-based research.

First, we have moved from a world where media organisations were gatekeepers to a world where media still create the news agenda, but platform companies control access to audiences

Established news media tend to be at the centre of online discussions of, for example, elections , and often drive the agenda .

But we have moved from a world where media organisations controlled both content and channels and we came to news directly by going to a specific broadcaster or publisher, to a world increasingly characterised by “distributed discovery”, where media organisations still create content, but people access it through platform channels like search engines, social media, and news aggregators.

In 2018, two-thirds of online news users surveyed across 37 different markets worldwide identified distributed forms of discovery as their main way of accessing and finding news online. (See Figure 1.) Amongst those under 35, three-quarters relied primarily on social media, search engines, and the like.

Second, this move to digital media and platforms generally does not generate filter bubbles but more diverse news diets

The automated serendipity of social media feeds and search engine results and incidental exposure (where people come across news while doing other things online) drive people to more and more diverse sources of information .

While echo chambers exist, where highly motivated minorities self-select into insular news diets and like-minded communities, fears of algorithmically generated filter bubbles currently seem misplaced.

Empirical research consistently finds that search engines and a wide range of different social media including both Facebook, Twitter, and YouTube demonstrably drive people to use more different sources of news, including more diverse sources and sources they do not seek out of their own volition. (See Figure 2.)

Third, journalism is often losing the battle for people’s attention and, in some countries, for the public’s trust

While a small minority of news lovers are extremely interested in news and access news several times a day, a far greater number of people access news less than once day.

Segmented on the basis of interest in news and frequency of access, our survey data from 37 markets shows that news lovers make up only 17% of the public, daily briefers about half (48%) and casual users, who access news less frequently than once a day, 35%. (See Figure 3.)

These patterns of news use points to a future of far greater information inequality - not because of lack of access, but because of limited interest , low engagement and, in some countries, little trust in news.

In Greece, just 26 percent say they trust most news most of the time, and in the United States 34 percent. But this is far from a universal phenomenon. Just north of the US border, 58 percent of Canadians say they trust most news, just like half the German population does.

And low levels of trust often obscure significant variation, especially along political lines. In the United States, just 17 percent of those on the political right say they trust most news, whereas half of those on the political left do. The differences in trust between the right and the left in the United States are bigger than the difference in trust between Greece and Germany.

Fourth, the business models that fund news are challenged, weakening professional journalism and leaving news media more vulnerable to commercial and political pressures

Even thirty years after the invention of the World Wide Web, the majority of professional journalism is still funded by newspapers . And an estimated 90% of publishers’ revenues worldwide still come from print , digital revenues are in many cases growing only slowly, and, where they exist, public service media are under considerable pressure.

Most of these existing forms of funding for professional journalism will decline as we continue to move to a more digital media environment where platforms like Google and Facebook capture most of the advertising, leading to further job cuts in newsrooms.

The risk here is not simply retrenchment and less coverage of many important issues, but also a less robust business of journalism more vulnerable to media capture by the state or politically motivated owners, and to pressure from advertisers.

Have you read?

Can you tell if this was written by a robot 7 challenges for ai in journalism, why credibility is the future of journalism, how data journalism can improve the world.

Fifth, news is more diverse than ever, and the best journalism in many cases better than ever, taking on everyone from the most powerful politicians to the biggest private companies.

It is clear that cost-cutting, increased pressure to produce more stories across more channels, and a 24/7 news cycle has led to a large volume of more superficial journalism.

At the same time, digital media have also allowed different marginalised voices to be heard and offer access to a far wider range of different sources and points of view.

And as journalists have embraced digital media and evolved various new formats, the best journalism in many countries today is in some ways better than ever – more independent of elite sources , more accessible, more timely, more informative, more interactive, more engaged with its audience.

The role of journalistic revelations in many different cases, including the #MeToo movement, in confronting corruption amongst public officials, and in fuelling public debate around platform companies’ power and privacy practices, underline the continued relevance of investigative reporting.

Where we are heading with digital media and journalism?

These five trends will not play out the same in every country. They will clearly differ depending on cultural, economic, political, and social context.

And with the continuing development of artificial intelligence, voice operated systems, and the integration of connected, digital media in more and more everyday objects, we are equally clearly only in the early phases of this fundamental change.

As we move forward, independent professional journalism will be more important than ever in helping people understand the major challenges and opportunities facing us, from day-to-day local events to global issues.

But as the business of news changes, journalism also risks becoming less robust, and ultimately incapable of helping the public make sense of our times or holding power to account.

This challenge is only compounded by increasingly open political hostility towards independent professional journalism, in the worst cases a veritable war on journalism .

In the absence of independent professional reporting providing accurate information, analysis, and interpretation, the public will increasingly rely on self-interested sources and rumours circulating online and offline, a shift that will hurt both the political process, civil society, and private enterprise.

At its best, independent professional journalism is essential for both the public good, politics, and private enterprise – and as it adapts to the digital media people all over the world are embracing, it can help ensure that this communications revolution results not in chaos, but in change for the better.

Rasmus Kleis Nielsen is Director of the Reuters Institute for the Study of Journalism and Professor of Political Communication at the University of Oxford.

Meera Selva is Director of the Journalist Fellowship Programme at the Reuters Institute for the Study of Journalism at the University of Oxford.

Don't miss any update on this topic

Create a free account and access your personalized content collection with our latest publications and analyses.

License and Republishing

World Economic Forum articles may be republished in accordance with the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International Public License, and in accordance with our Terms of Use.

The views expressed in this article are those of the author alone and not the World Economic Forum.

The Agenda .chakra .wef-n7bacu{margin-top:16px;margin-bottom:16px;line-height:1.388;font-weight:400;} Weekly

A weekly update of the most important issues driving the global agenda

• Ethics & Leadership
• Fact-Checking
• Media Literacy
• The Craig Newmark Center
• Reporting & Editing
• Ethics & Trust
• Tech & Tools
• Business & Work
• Educators & Students
• Training Catalog
• Custom Teaching
• For ACES Members
• All Categories
• Broadcast & Visual Journalism
• Fact-Checking & Media Literacy
• In-newsroom
• Memphis, Tenn.
• Minneapolis, Minn.
• St. Petersburg, Fla.
• Washington, D.C.
• Poynter ACES Introductory Certificate in Editing
• Poynter ACES Intermediate Certificate in Editing
• Ethics & Trust Articles
• Get Ethics Advice
• Fact-Checking Articles
• IFCN Grants
• International Fact-Checking Day
• Teen Fact-Checking Network
• International
• Media Literacy Training
• MediaWise Resources
• Ambassadors
• MediaWise in the News

Support responsible news and fact-based information today!

Opinion | How AI will reshape journalism’s narrative in 2024

As ai transforms the media landscape, a blend of technology and human touch will be the key to success..

In the intricate dance of media evolution, 2024 marks a pivotal moment, when generative artificial intelligence steps into the spotlight, simultaneously leading and following in the rhythm of innovation.

This dance is not new. It echoes past transformations in social media where pioneers like BuzzFeed and Vice once led, challenging legacy media to adapt or fall behind.

The music, however, has changed, and with it, the steps. Here are three things we should keep in mind to start 2024.

The pendulum swings: paywalls and niche subscriptions

The business interests of media organizations are in a constant ebb-and-flow. The current dominant waltz has seen the steady rise of paywalls and regards subscription-based models as having ultimate value. This trend is a reaction to the social media years, with its pivots to video and audience attention as the ultimate key performance indicator.

As Sam Cholke, the manager of distribution and audience growth at the Institute for Nonprofit News, writes , “The last year has reminded many news organizations how brittle online audiences can be. Facebook traffic declined sharply in 2023, dashing publishers’ hopes that a long fall in referral traffic had finally plateaued.”

The social media epoch brought Vice, Mic, BuzzFeed, and others who promised investors they would master all social media trends and elbow out the competition for the kind of scale hitherto unimaginable to media companies. We’ve seen the results of that overreach. Eventually, a platform comes along that requires too much retooling, or monetization needs to happen further down the business funnel (such as subscriptions) and business leaders realize short-term audience attention isn’t fit for a cause. This cycle will rinse and repeat.

Now, as the pendulum swings towards exclusivity, a gap will eventually emerge for media organizations to again leverage mass audiences for a business advantage. However, this return to a mass audience won’t happen for a few years. The social years from 2008 to 2016 gave way to our current era. When the tide does turn again back to upstarts making a name for themselves by amassing large audiences it will look different from the aughts, as the technological emphasis will shift from social to generative AI.

The rarity of the human touch: AI-generated vs. unique content

As generative AI takes the stage in the production of content — from punchy headlines to tailored audience segments — it inadvertently elevates the value of human-created content.

I wrote about this in Poynter as a “ proof of trust ” premium that will be associated with provably bylined or printed content. Media analyst Mauricio Cabrera and others are also coming to this conclusion. “It might sound paradoxical,” Cabrera wrote, “but the future will require the media to be like magazines and print newspapers tend to be: with less content, but of better quality.”

Technology leader Ben Werdmuller used similar terms: “Newsrooms that commit to AI-driven storytelling as a way to cut costs while increasing output will be lost in a sea of similarly bland content and spammy marketing. Newsrooms that cling to traditional SEO and social media tactics will find that they become less and less effective in the face of more and more noise.”

The uniqueness of personally crafted bylines and narratives will become akin to rare art in a world of replicas. This will not diminish AI’s role in the future of content production, but rather complement it. Generative AI is not a panacea for content, but a new paradigm that will put value and emphasis on the human touch.

The next era of engagement journalism

Engagement journalism, a branch of journalism with a long and evolving tradition, will also find new life through generative AI. At its most basic, engagement journalism is about making the process of journalism more transparent and participatory.

While the first blush of generative AI applications is around producing traditional articles, it can also help to sift through public input, separating the wheat from the chaff, and revive old engagement tools like comment sections and tip lines.

Even more interesting would be the scalability of an AI editor that can craft narratives from a mosaic of public input. A question such as, “What do you think about the mayor’s proposal?” can go from “leave a comment” to “see what the public thinks en-masse” with relative ease. This evolution speaks not of AI as a creator in solitude, but as a collaborator and community builder, fostering deeper connections between the media and its audience.

In 2024, we stand at the cusp of a media renaissance, in which AI’s role is not just as a tool but as a partner in the journalistic process. This partnership calls for a delicate balance, ensuring that while we embrace the efficiencies and insights AI offers, we also preserve the irreplaceable value of human perspective and creativity.

As we navigate these waters, questions arise: How do we best integrate AI without losing the soul of journalism? Can AI be taught to understand the nuances of human experience that are often the heart of impactful storytelling? And most importantly, how do we maintain the trust of our audience in an era where the line between machine-generated and human-crafted content blurs?

Opinion | Inside Sally Buzbee’s departure and what’s next for The Washington Post

Buzbee’s abrupt resignation and the ensuing plans have staffers angry, confused and curious about the future of one of America’s great institutions

Opinion | AP Stylebook’s new chapter on crime is a glimpse into the future

A decade from now, American newsrooms will have replaced cheap stories with data-rich narratives that educate communities and hold cops accountable

Does Trump’s felony conviction bar him from owning a gun?

Federal law prohibits people with felony convictions from possessing or acquiring firearms

Opinion | Sally Buzbee steps down as executive editor of The Washington Post

Buzbee joined the Post in 2021 after a career at The Associated Press. She was the first female editor of the Post, which was founded in 1877.

What do horse race journalists think of ‘horse race journalism’?

As the 2024 presidential election approaches, breathless reporting of incremental polling has already begun.

Start your day informed and inspired.

Get the Poynter newsletter that's right for you.

• Search Menu
• Sign in through your institution
• Advance Articles
• Author Guidelines
• Submission Site
• Open Access
• Why Submit?
• About Journal of Communication
• About International Communication Association
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

• < Previous

Remaking the news: Essays on the future of journalism scholarship in the digital age

• Article contents
• Figures & tables
• Supplementary Data

James S McLean, Remaking the news: Essays on the future of journalism scholarship in the digital age, Journal of Communication , Volume 68, Issue 3, June 2018, Pages E35–E37, https://doi.org/10.1093/joc/jqy020

• Permissions Icon Permissions

As editors Pablo Boczkowski and C. W. Anderson explain in the introduction to Remaking the News , this is a collection of essays “about the relationship between technology and journalism.” The “technology” is, of course, digital and the “journalism” is pretty much everything that digital technologies touch—both in production and circulation—across a vast landscape of computer-enabled reportage, from the trivial to the profound. What an ambitious undertaking!

Yet the ambition is not misplaced. There is serious scholarship here, organized across four sections that embrace and interrogate broad categories: theory and methodology; key concepts; occupational culture and practice; and underexamined themes. Furthermore, unlike many collected works, Remaking the News does not suffer from inconsistency, wherein one or two disappointing contributions tend to spoil the effect of the whole. On the contrary, this is an extraordinarily even presentation where the reader can be assured of taking away something of value from each essay on offer. And should the trajectory of a section become somehow muddled there are helpful commentaries by senior scholars to help us resituate thematic purpose.

Email alerts

Citing articles via.

• Recommend to your Library

Affiliations

• Online ISSN 1460-2466
• Print ISSN 0021-9916
• Copyright © 2024 International Communication Association
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Remaking the News : Essays on the Future of Journalism Scholarship in the Digital Age

Pablo J. Boczkowski is Hamad Bin Khalifa Al-Thani Professor in the Department of Communication Studies at Northwestern University.

C. W. Anderson is Associate Professor at the College of Staten Island (CUNY); as of the Fall of 2017 he will be a Professor of Media and Communication at the University of Leeds. He is the author, coauthor, or coeditor of multiple books and articles on digital journalism, sociology, political communication, and science and technology studies.

Leading scholars chart the future of studies on technology and journalism in the digital age.

The use of digital technology has transformed the way news is produced, distributed, and received. Just as media organizations and journalists have realized that technology is a central and indispensable part of their enterprise, scholars of journalism have shifted their focus to the role of technology. In Remaking the News , leading scholars chart the future of studies on technology and journalism in the digital age.

These ongoing changes in journalism invite scholars to rethink how they approach this dynamic field of inquiry. The contributors consider theoretical and methodological issues; concepts from the social science canon that can help make sense of journalism; the occupational culture and practice of journalism; and major gaps in current scholarship on the news: analyses of inequality, history, and failure.

Contributors Mike Ananny, C. W. Anderson, Rodney Benson, Pablo J. Boczkowski, Michael X. Delli Carpini, Mark Deuze, William H. Dutton, Matthew Hindman, Seth C. Lewis, Eugenia Mitchelstein, W. Russell Neuman, Rasmus Kleis Nielsen, Zizi Papacharissi, Victor Pickard, Mirjam Prenger, Sue Robinson, Michael Schudson, Jane B. Singer, Natalie (Talia) Jomini Stroud, Karin Wahl-Jorgensen, Rodrigo Zamith

• Permissions
• Cite Icon Cite

Remaking the News : Essays on the Future of Journalism Scholarship in the Digital Age Edited by: Pablo J. Boczkowski, C. W. Anderson https://doi.org/10.7551/mitpress/10648.001.0001 ISBN (electronic): 9780262339681 Publisher: The MIT Press Published: 2017

Download citation file:

• Ris (Zotero)
• Reference Manager

Table of Contents

• [ Front Matter ] Doi: https://doi.org/10.7551/mitpress/10648.003.0030 Open the PDF Link PDF for [ Front Matter ] in another window
• Acknowledgments Doi: https://doi.org/10.7551/mitpress/10648.003.0002 Open the PDF Link PDF for Acknowledgments in another window
• Contents Doi: https://doi.org/10.7551/mitpress/10648.003.0001 Open the PDF Link PDF for Contents in another window
• Introduction: Words and Things By Pablo J. Boczkowski , Pablo J. Boczkowski Search for other works by this author on: This Site Google Scholar C. W. Anderson C. W. Anderson Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0003 Open the PDF Link PDF for Introduction: Words and Things in another window
• 1: Scholarship on Online Journalism: Roads Traveled and Pathways Ahead By Pablo J. Boczkowski , Pablo J. Boczkowski Search for other works by this author on: This Site Google Scholar Eugenia Mitchelstein Eugenia Mitchelstein Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0005 Open the PDF Link PDF for 1: Scholarship on Online Journalism: Roads Traveled and Pathways Ahead in another window
• 2: From Heterogeneity to Differentiation: Searching for a Good Explanation in a New Descriptivist Era By Rodney Benson Rodney Benson Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0006 Open the PDF Link PDF for 2: From Heterogeneity to Differentiation: Searching for a Good Explanation in a New Descriptivist Era in another window
• 3: Rediscovering the News: Journalism Studies’ Three Blind Spots By Victor Pickard Victor Pickard Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0007 Open the PDF Link PDF for 3: Rediscovering the News: Journalism Studies’ Three Blind Spots in another window
• 4: Newsroom Ethnography and Historical Context By C. W. Anderson C. W. Anderson Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0008 Open the PDF Link PDF for 4: Newsroom Ethnography and Historical Context in another window
• Commentary: Reflections on Scholarship in the Study of Online News By William H. Dutton William H. Dutton Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0009 Open the PDF Link PDF for Commentary: Reflections on Scholarship in the Study of Online News in another window
• 5: Digital News as Forms of Knowledge: A New Chapter in the Sociology of Knowledge By Rasmus Kleis Nielsen Rasmus Kleis Nielsen Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0011 Open the PDF Link PDF for 5: Digital News as Forms of Knowledge: A New Chapter in the Sociology of Knowledge in another window
• 6: On the Worlds of Journalism By Seth C. Lewis , Seth C. Lewis Search for other works by this author on: This Site Google Scholar Rodrigo Zamith Rodrigo Zamith Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0012 Open the PDF Link PDF for 6: On the Worlds of Journalism in another window
• 7: The Whitespace Press: Designing Meaningful Absences into Networked News By Mike Ananny Mike Ananny Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0013 Open the PDF Link PDF for 7: The Whitespace Press: Designing Meaningful Absences into Networked News in another window
• Commentary: Remaking Events, Storytelling, and the News By Zizi Papacharissi Zizi Papacharissi Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0014 Open the PDF Link PDF for Commentary: Remaking Events, Storytelling, and the News in another window
• 8: Helping Newsrooms Work toward Their Democratic and Business Objectives By Natalie Jomini Stroud Natalie Jomini Stroud Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0016 Open the PDF Link PDF for 8: Helping Newsrooms Work toward Their Democratic and Business Objectives in another window
• 9: Journalism Ethics and Digital Audience Data By Matthew Hindman Matthew Hindman Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0017 Open the PDF Link PDF for 9: Journalism Ethics and Digital Audience Data in another window
• 10: Reinventing Journalism as an Entrepreneurial Enterprise By Jane B. Singer Jane B. Singer Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0018 Open the PDF Link PDF for 10: Reinventing Journalism as an Entrepreneurial Enterprise in another window
• Commentary: Blurring Boundaries By W. Russell Neuman W. Russell Neuman Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0019 Open the PDF Link PDF for Commentary: Blurring Boundaries in another window
• 11: Check Out This Blog: Researching Power and Privilege in Emergent Journalistic Authorities By Sue Robinson Sue Robinson Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0021 Open the PDF Link PDF for 11: Check Out This Blog: Researching Power and Privilege in Emergent Journalistic Authorities in another window
• 12: A History of Innovation and Entrepreneurialism in Journalism By Mirjam Prenger , Mirjam Prenger Search for other works by this author on: This Site Google Scholar Mark Deuze Mark Deuze Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0022 Open the PDF Link PDF for 12: A History of Innovation and Entrepreneurialism in Journalism in another window
• 13: A Manifesto of Failure for Digital Journalism By Karin Wahl-Jorgensen Karin Wahl-Jorgensen Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0023 Open the PDF Link PDF for 13: A Manifesto of Failure for Digital Journalism in another window
• Commentary: The Journalism Studies Tree By Michael Schudson Michael Schudson Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0024 Open the PDF Link PDF for Commentary: The Journalism Studies Tree in another window
• Postscript: The Who, What, When, Where, Why, and How of Journalism and Journalism Studies By Michael X. Delli Carpini Michael X. Delli Carpini Search for other works by this author on: This Site Google Scholar Doi: https://doi.org/10.7551/mitpress/10648.003.0025 Open the PDF Link PDF for Postscript: The Who, What, When, Where, Why, and How of Journalism and Journalism Studies in another window
• Contributors Doi: https://doi.org/10.7551/mitpress/10648.003.0026 Open the PDF Link PDF for Contributors in another window
• References Doi: https://doi.org/10.7551/mitpress/10648.003.0027 Open the PDF Link PDF for References in another window
• Index Doi: https://doi.org/10.7551/mitpress/10648.003.0028 Open the PDF Link PDF for Index in another window
• Inside Technology Doi: https://doi.org/10.7551/mitpress/10648.003.0029 Open the PDF Link PDF for Inside Technology in another window
• Open Access

A product of The MIT Press

Mit press direct.

• About MIT Press Direct

Information

• Accessibility
• For Authors
• For Customers
• For Librarians
• Direct to Open
• Media Inquiries
• Rights and Permissions
• For Advertisers
• About the MIT Press
• The MIT Press Reader
• MIT Press Blog
• Seasonal Catalogs
• MIT Press Home
• Give to the MIT Press
• Direct Service Desk
• Terms of Use
• Privacy Statement
• Crossref Member
• COUNTER Member  
• The MIT Press colophon is registered in the U.S. Patent and Trademark Office

This Feature Is Available To Subscribers Only

Sign In or Create an Account

• Close Menu Search
• Tips and Lessons
• Classroom in a Box
• Journalism Training
• News Literacy Principles
• Media Literacy Articles
• Curriculum and Lessons
• Q and A with the Pros

SchoolJournalism.org

Future of journalism.

An essay on the future of journalism by Abbey Dean, Missouri School of Journalism graduate student

I understand what you’re thinking. No really, I do. Like you, I’ve been there before. I was a young would-be journalist who wasn’t sure if I was entering the right field. My parents understood my passion, but I could see the glimmer of anxiety in their eyes when I officially decided to pursue journalism.

The future is scary enough without the doom and gloom so often cast upon the field of journalism. But I’m here to tell you that the truth is never as disparaging once you ask the right people.

What naysayers fail to do in spinning the future of journalism negatively is to separate the future of print journalism from the overarching field of journalism. I promise; there’s a big difference. Print publications are having a rough go of it, that’s true; however, there are still a plethora of opportunities for young aspiring journalists in print and digital journalism.

Consider how you’re reading this right now. On your phone? iPad? Maybe you even found the link to this article through our Twitter feed? With the proliferation of the digital age and social media technology, public demand has driven the desire for 24/7 access for news to record heights. For many publications, the goal now is to figure out how best to combine the traditional techniques of journalism with new, digital platforms. With this audience demand must come suppliers, and that’s where you come in.

So, what does it take to land your dream job or any job in the industry? Aside from a boundless curiosity, solid writing skills and passion for the field, you also need the flexibility and know-how to work with changing technologies. Nowadays, a working knowledge of video, audio and Web technology is required for most news reporters. In short, you need to be a Jack or Jill of all trades. Take a look at this sample job description for a newsroom position at The New York Times .

But studying journalism doesn’t just teach you how to be a good journalist. I’ve learned priceless skills that could easily be transferable to other fields. To start, you will learn how to write, write well and write on deadline; there isn’t an employer out there who wouldn’t appreciate this. You will also learn organizational skills and how to communicate with people from all walks of life, which will serve you well throughout any career you choose to undertake. Not to mention, all of the above are incredibly important life skills that are valuable whether you decide journalism is for you or not.

The field of journalism is a tumultuous one that experiences ups and downs like most other areas of work. But with these rises and declines also come new, exciting possibilities within the industry. Whether you’re interested in anything from investigative reporting to magazine editing, be reassured that there is a job out there to match your interests.

Remember that journalism will always be around, too–that’s for certain. How we choose to tell and frame our stories will continue to evolve, but the values that drive good journalism won’t. Strong journalistic ethics, transparency in reporting and an unflinching devotion to the truth will always spur the best journalism and attract the most readers. With these skills and values, you can become the next generation of muckrakers and digital gurus with all the old-school pizazz of Woodward and Bernstein.

Still not convinced that the future of journalism is quite this bright? Take a glance at some of the articles below to see what some of the industry’s top leaders are saying and get tips on how to prepare for your future.

Tom Rosenstiel, executive director of the American Press Institute, dispels five myths about the future of journalism .

Arianna Huffington, chair, president and editor-in-chief of the Huffington Post Media Group, discusses her vision of journalism’s future .

The Poynter Institute conducted a study on the future of journalism education . The results might surprise you.

The Poynter Institute has also initiated conversation about how to modernize journalism education for the next generation of journalists.

Discover 10 tips designed to prep aspiring young journalists on the road ahead from Forbes.com.

The Poynter Institute offers more tips and tidbits on how to become a journalist in the digital age.

Learn about a new, emerging form of journalism that could be a match for you from this article at Slate.com.

Huff Post College takes a look into how social media is changing the digital landscape .

• Fellowships
• Business Models
• Mobile & Apps
• Audience & Social
• Aggregation & Discovery
• Reporting & Production
• Translations

What’s with the rise of “fact-based journalism”?

Here’s a term you may be hearing with increasing frequency: “Fact-based journalism.” The Associated Press uses it in fund-raising appeals , as does ProPublica , and our local NPR affiliate. The National Association of Broadcasters and the Corporation for Public Broadcasting both describe themselves as purveyors of “fact-based journalism” in their public relations materials.

Even news outlets with an overtly partisan bent employ some variation of the term. The right-leaning The Dispatch , for instance, describes itself as a source of “fact-based conservative news.” The U.S. Agency for International Development uses the term as a guiding concept for its media development work.

When and why did this term rise to prominence? We did a keyword search of “fact-based journalism” in NewsBank , a news repository of over 12,000 sources, for the years 1990 through 2023.

As the graph below indicates, usage of the term ballooned starting in 2016 and saw a big spike in 2021. And as the graph also indicates, the term “fact-based journalism” was rarely used prior to the early 2000s.

The increasing usage of the term corresponds with the beginning of Trump’s presidency. Given this timing, we next conducted a parallel search of the term “fake news.” Our suspicion was that the term “fact-based journalism” arose in response to the rise of the notion of “fake news” that so dominated the discourse around journalism and politics during the Trump presidency . The results support our hypothesis.

As the next graph indicates, usage of the term “fact-based journalism” began escalating right as the term “fake news” took off in popularity. Though correlation doesn’t guarantee causality, the pattern below strongly suggests that the term’s use in the news industry may have been a response to the prevalence of “fake news” discourse within politics and the media.

As we noted above, usage of the term “fact-based journalism” experienced a massive spike in usage in 2021, when Trump was no longer president and usage of the term “fake news” had subsided. An examination of stories published in 2021 suggests why: That year, the Nobel Committee awarded its Peace Prize to two journalists, Maria Ressa and Dmitry Muratov, who pushed back against authoritarian regimes. The Nobel Committee hailed their work as indicative of the power of “free, independent, and fact-based journalism,” and this phrase appeared in multiple stories about the awarding of the prize.

As the graphs indicate, the use of “fact-based journalism” has remained relatively high in the years since the Nobel announcement, suggesting that the phrase has become an established part of the vocabulary around journalism.

As our politics and media have grown more polarized, and as the barriers to entry to operating a “news” organization evaporated, the media ecosystem has become a breeding ground for misinformation posing as news . In such an environment, identifying some approaches to journalism, or some news organizations, as bastions of “fact-based journalism” has become a way of trying to distinguish legitimate journalism from the rest of the dreck.

Similarly, as the Trump era ushered in the strategy of politicians slandering legitimate reporting as “fake news,” self-identifying as “fact-based journalism” became a way for news organizations to push back against efforts by politicians or hyperpartisan media outlets to discredit them.

However, while there may be compelling reasons to embrace the term “fact-based journalism,” doing so represents a damaging concession. To describe one form of journalism as “fact-based” is to tacitly acknowledge that there is also such a thing as “non-fact-based journalism.” And there isn’t.

“Fact-based journalism” is what linguists call a pleonasm — a redundancy in linguistic expression, like “black darkness.” All journalism is fact-based. If it’s not, then it’s not journalism.

Sure, opinion and partisanship have long played prominent roles in journalism . But for opinion to merit the label “opinion journalism ,” or for partisan reporting to merit being called “partisan journalism ,” they still need to be grounded in fact. (Interpretation of those shared facts can, of course, differ.)

The widespread embrace of the term “fact-based journalism” within the journalism community is a capitulation to those who are undermining the very notion of journalism, and thereby destabilizing the foundations of our democracy. We need to do more to defend and evangelize the parameters of journalism, not cede linguistic space to those seeking to blur these parameters as part of a broader political strategy.

If some sort of qualifier is indeed necessary in this age of widespread efforts to disguise falsity and political influence operations as journalism, then how about a term that goes on the offensive — perhaps “legitimate journalism” or “authentic journalism”? Such terms better distinguish organizations that are producing journalism from those that are making a mockery of it.

Philip M. Napoli is the James R. Shepley Professor of Public Policy, in the Sanford School of Public Policy at Duke University, where he is also the director of the DeWitt Wallace Center for Media & Democracy. Asa Royal is an associate in research in the DeWitt Wallace Center for Media & Democracy, where he leads research initiatives on topics such as the economics and impact of local journalism, and the rise of “pink slime” news networks.

Cite this article Hide citations

Napoli, Philip M.. "What’s with the rise of “fact-based journalism”?." Nieman Journalism Lab . Nieman Foundation for Journalism at Harvard, 29 May. 2024. Web. 4 Jun. 2024.

Napoli, P. (2024, May. 29). What’s with the rise of “fact-based journalism”?. Nieman Journalism Lab . Retrieved June 4, 2024, from https://www.niemanlab.org/2024/05/whats-with-the-rise-of-fact-based-journalism/

Napoli, Philip M.. "What’s with the rise of “fact-based journalism”?." Nieman Journalism Lab . Last modified May 29, 2024. Accessed June 4, 2024. https://www.niemanlab.org/2024/05/whats-with-the-rise-of-fact-based-journalism/.

{{cite web     | url = https://www.niemanlab.org/2024/05/whats-with-the-rise-of-fact-based-journalism/     | title = What’s with the rise of “fact-based journalism”?     | last = Napoli     | first = Philip M.     | work = [[Nieman Journalism Lab]]     | date = 29 May 2024     | accessdate = 4 June 2024     | ref = {{harvid|Napoli|2024}} }}

To promote and elevate the standards of journalism

Covering thought leadership in journalism

Pushing to the future of journalism

Exploring the art and craft of story

The Nieman Journalism Lab is a collaborative attempt to figure out how quality journalism can survive and thrive in the Internet age.

It’s a project of the Nieman Foundation for Journalism at Harvard University .

• Subscribe to our email
• Follow us on Twitter
• Like us on Facebook
• Download our iPhone app
• Subscribe via RSS
• Tweet archive

Steve Karnowski, Associated Press Steve Karnowski, Associated Press

Leave your feedback

• Copy URL https://www.pbs.org/newshour/nation/bag-of-120000-left-as-bribe-for-juror-in-minnesota-food-program-fraud-trial

Bag of $120,000 left as bribe for juror in Minnesota food program fraud trial

MINNEAPOLIS (AP) — Federal authorities in Minnesota have confiscated cellphones and taken all seven defendants into custody as investigators try to determine who attempted to bribe a juror with a bag of cash containing $120,000 to get her to acquit them on charges of stealing more than $40 million from a program meant to feed children during the pandemic.

The case went to the jury late Monday afternoon, after the juror, who promptly reported the attempted bribe to police, was dismissed and replaced with an alternate. The incident had further ripple effects before deliberations resumed Tuesday — when another juror was replaced after a family member asked about the the attempted bribe.

READ MORE: Federal authorities charge 47 with stealing $250 million from food program for low-income children

According to an FBI agent’s affidavit, a woman rang the doorbell at the home of “Juror #52” in the Minneapolis suburb of Spring Lake Park late Sunday, the night before the case went to the jury. The juror wasn’t home, but a relative answered the door. The woman handed the relative a gift bag with a curly ribbon and images of flowers and butterflies and said it was a “present” for the juror.

“The woman told the relative to tell Juror #52 to say not guilty tomorrow and there would be more of that present tomorrow,” the agent wrote. “After the woman left, the relative looked in the gift bag and saw it contained a substantial amount of cash.”

The juror called police right after she got home and gave them the bag of cash. It held $100, $50 and $20 bills totaling around $120,000. The FBI took the bag from Spring Lake Park police on Monday morning and interviewed the juror.

The woman who left the bag knew the juror’s first name, the agent said. Names of the jurors have not been made public, but the list of people who had access to it included prosecutors and defense lawyers —- and the seven defendants themselves.

READ MORE: Justice Department announces charges against hundreds of alleged COVID-19 fraudsters

U.S. District Judge Nancy Brasel and attorneys for both sides learned about the attempted bribe Monday morning. The judge ordered all seven defendants to surrender their cellphones at the request of the government so that investigators could look for evidence. She also ordered all seven taken into custody.

“It is highly likely that someone with access to the juror’s personal information was conspiring with, at a minimum, the woman who delivered the $120,000 bribe,” the FBI agent wrote, noting that the alleged fraud conspiracy at the heart of the trial involved electronic communications, including text messages and emails.

Before the case went to the jury late Monday afternoon, Brasel ordered them sequestered for deliberations. When one of them called home to say she’d been sequestered, according to KARE-TV and KSTP-TV, a family member asked, “Is it because of the bribe?” The judge replaced that juror with an alternate, too.

Anyone involved in the attempted bribe could face federal charges of bribery of a juror and influencing a juror, with a maximum potential penalty of 15 years in prison.

Minneapolis FBI spokesperson Diana Freedman said Tuesday that she could not provide information about the ongoing investigation.

According to the Star Tribune, Assistant U.S. Attorney Joseph Thompson told the court Monday: “This is completely beyond the pale. This is outrageous behavior. This is stuff that happens in mob movies.” Defense attorney Andrew Birrell called it “a troubling and upsetting accusation,” according to the newspaper.

The seven were the first of 70 defendants to go on trial in what federal prosecutors have called one of the largest COVID-19-related fraud cases in the country. They’ve described it as a massive scheme to exploit lax rules during the pandemic and steal from a program that was meant to provide meals to children in Minnesota.

Prosecutors have said the seven collectively stole over $40 million in a conspiracy that cost taxpayers $250 million. At the center of the alleged plot was a group called Feeding Our Future. Prosecutors say just a fraction of the money went to feed low-income kids, and that the rest was spent on luxury cars, jewelry, travel and property. Federal authorities say they have recovered about $50 million.

Eighteen other defendants have already pleaded guilty, while the rest are awaiting trial. Among them is Aimee Bock, the founder of Feeding our Future. She has maintained her innocence, saying she never stole and saw no evidence of fraud among her subcontractors.

The defendants are: Abdiaziz Shafii Farah, Mohamed Jama Ismail, Abdimajid Mohamed Nur, Said Shafii Farah, Abdiwahab Maalim Aftin, Mukhtar Mohamed Shariff and Hayat Mohamed Nur. The charges against them include wire fraud and money laundering. Shariff was the only defendant to testify and the only one to call witnesses on his behalf.

The food aid came from the U.S. Department of Agriculture and was administered by the state. Nonprofits and other partners under the program were supposed to serve meals to kids. Defendants allegedly produced invoices for meals that were never served, ran shell companies, laundered money, indulged in passport fraud, and accepted kickbacks.

An Associated Press analysis published last June documented how thieves across the country plundered billions in federal COVID-19 relief dollars. Fraudsters potentially stole more than $280 billion, while another $123 billion was wasted or misspent. Combined, the loss represented 10% of the $4.3 trillion the government disbursed in COVID relief by last fall. Nearly 3,200 defendants have been charged, according to the U.S. Justice Department. About $1.4 billion in stolen pandemic aid has been seized.

Support Provided By: Learn more

Educate your inbox

Subscribe to Here’s the Deal, our politics newsletter for analysis you won’t find anywhere else.

Thank you. Please check your inbox to confirm.

June 4, 2024

The 180-Year-Old Endnotes That Foretold the Future of Computation

Ada Lovelace’s wisdom about the first general-purpose computer can be found buried in the appendix of another paper

By Jack Murtagh

This diagram shows an algorithm for the analytical engine for the computation of Bernoulli numbers, taken from the article Sketch of the Analytical Engine Invented by Charles Babbage, written by Luigi Menabrea with notes by Ada Lovelace.

The Picture Art Collection/Alamy Stock Photo

Many fields of science can point to a foundational document: Isaac Newton’s Principia for the physics of classical mechanics or Charles Darwin’s On the Origin of Species for evolutionary biology. But only computer science can claim its foundational document hides in the endnotes.

Augusta Ada King, Countess of Lovelace, better known as Ada Lovelace , was commissioned in 1842 to translate a paper about the world’s first general-purpose computer. She appended her own annotations, which ran three times longer than the original article and completely eclipsed it in terms of technical meat and philosophical insights. With impressive foresight, they established her as the first person to envision the universal capabilities of computers that we take for granted today.

Though the whole world knows of Lovelace’s dad, she never knew him personally. Lord Byron, celebrated for his Romantic poetry, was by all accounts a rotten husband and absent father . (After first meeting Lord Byron in 1812, aristocrat Lady Caroline Lamb apparently called him “ mad, bad and dangerous to know .”) Lovelace’s parents’ brief and tumultuous marriage ended with Lady Byron, née Annabella Milbanke, accusing him of abuse and infidelity a mere five weeks after his daughter’s birth. Although they never met again, Lord Byron had at least two lasting effects on his daughter. For one, she maintained a curiosity about him and his work and perceived much of her later scientific worldview through a poetic lens. Second, Lady Byron believed that she could shield Lovelace from inheriting her father’s erratic temperament by steering her away from literary studies and instead fostering her interest in science and math .

On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing . By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.

The defining partnership of Lovelace’s intellectual life would come through polymath Charles Babbage, whom she met at age 17. In 1833, the year that they met, Babbage had begun designing the first general-purpose mechanical computer, called the analytical engine . The idea captivated Lovelace, and the two became close friends and collaborators.

Had he ever managed to build it, Babbage’s analytical engine would have been a marvel to behold. Sketches and spec sheets depict an intricate beast with rattling gear columns towering 15 feet high, lever panels cranking in lockstep and thousands of moving parts churning together in a sort of steampunk orchestra. Although unrecognizable as your laptop’s ancestor, the device remarkably contained most of the same components that we know in modern computer architecture. It had a central processing unit called a mill to carry out arithmetic operations, memory storage, input capabilities to read data and programs, and even a printer.

Ada Lovelace, shown here in this portrait, is sometimes called the world's first computer programmer.

API/Gamma-Rapho via Getty Images

Keep in mind that the era predated electronics, so steam would have powered all of this functionality. Every gear column stacked 40 gears with 10 teeth each labeled 0 through 9 for storing a 40-digit number, much like the disks on a combination lock. The mill could crunch these numbers through complicated procedures of rotating and swapping gears. But the secret sauce that separated the analytical engine from its predecessors was that it was programmable.

The design drew inspiration from the Jacquard machine , which attached to a loom and disrupted the 19th-century textile industry by automating weaving through punch card technology. Like the Jacquard machine, the analytical engine could have read instructions in the form of holes punched into card stock. Babbage had built a prototype for an earlier device called the difference engine (currently on display at the Museum of Science in London ), which could mechanically calculate a preset suite of operations such as certain logarithms and trigonometric functions. He abandoned it because he saw greater potential in a machine that could be programmed to perform arbitrary calculations.

Programmability elevates a machine from a mere calculator to a computer. The analytical engine would have been able to choose which instruction to follow based on the outcome of a previous computation—a skill called conditional branching. This capability would have allowed the machine to carry out advanced instructions such as if-else statements and loops seen in today’s programming languages. Although Babbage understood the power of programmability, he still viewed the analytical engine as a purely mathematical device . Only Lovelace foresaw the true potential in the computer.

Ten years after Lovelace first learned of the analytical engine, she was commissioned to translate a paper on the topic written by mathematician and engineer Luigi Federico Menabrea (who would go on to be Italy’s seventh prime minister). Possessing much more detailed knowledge of the engine than Menabrea, Lovelace corrected his errors and added seven of her own endnotes , which alone constitute a watershed document in the history of computation.

Many retrospectives focus on “ Note G ” because it contains the first published computer program. Lovelace’s program calculates Bernoulli numbers , which play a critical role in a branch of math called analysis. Her calculation method used more computational steps than necessary, but she chose this tack deliberately to show off the power and flexibility of the machine. Although many credit her as the first computer programmer, some historians contend that unpublished programs predated her notes and that technically Note G doesn’t contain a program as we’d name it today but rather an execution trace —a record of every operation performed during the execution of a program. I find such quibbles moot because I’d argue that the moniker “first computer programmer” undersells the wisdom found in the rest of her notes.

Babbage saw the analytical engine as a mathematical device. After all, it primarily stored and operated on numbers. But Lovelace recognized that a machine designed to crunch numbers could do much more if the numbers represented other things . For example, she wrote in the endnotes, “Supposing, for instance, that the fundamental relations of pitched sounds in the science of harmony and of musical composition were susceptible of such expression and adaptations, the engine might compose elaborate and scientific pieces of music of any degree of complexity or extent.” Now, some 180 years later, generative artificial intelligence tools such as Suno and Udio can compose music from text-based prompts.

Lovelace’s insight marks a profound conceptual leap that wouldn’t be formalized for another century with the work of Alan Turing : Fundamentally, computation involves the manipulation of symbols according to rules. There’s no limit on what those symbols represent. This idea is built into Turing’s mathematical model of computation, and it originated with Lovelace. We take for granted today that the same bits of 0’s and 1’s encode every type of media—text, images, audio, video—but it’s hard to fathom that this future had been envisioned before the first computer had even been built.

Lovelace also explicitly discussed artificial intelligence, kicking off a debate that defines the modern era. She concluded in her endnotes that the analytical engine would not be intelligent because it wouldn’t originate anything, arguing that “it can do whatever we know how to order it to perform.” Turing challenged Lovelace by name in his seminal paper on “thinking machines.” Though he conceded that computers merely do what they’re told, Turing suggested they still have the power to surprise us. Turing acknowledged that Lovelace never had an opportunity to interact with a computer and so lacked the advantage of experiencing such surprise for herself. In today’s AI landscape, many now believe that machines can exhibit intelligence (although holdouts from Lovelace’s camp are not hard to find ). The fact that Lovelace even entertained the question of whether a clanging heap of cogs that had only been sketched on paper would be intelligent shows how ahead of her time she was.

Ultimately Babbage’s contentious relationship with the British government meant that he never secured enough funding to bring the analytical engine to life. It’s funny to contrast Lovelace’s prophetic words about the potential of computers with a quote from the U.K.’s then prime minister Robert Peel : “What shall we do to get rid of Mr. Babbage and his calculating machine? Surely if completed it would be worthless as far as science is concerned?”

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

Enter the email address you signed up with and we'll email you a reset link.

• We're Hiring!
• Help Center

The Future of Journalism

2016, Journalism Studies

Related Papers

Journalism Practice

Cynthia Carter

Blanca castro

Barbie Zelizer

María Luengo

Jari Kaivo-oja

George Brock

Antxoka Agirre

This work collates the opinions of European professionals and internet users concerning the changing reality facing the world of journalism following the explosion of digital media. Our research shows the results of analytical surveys conducted between October 2013 and January 2014 among 54 professionals of renowned prestige, and 500 regular users of the digital media in the five most populated countries in the European Union: Germany, United Kingdom, France, Italy and Spain, all states affected to a greater or lesser extent by the crisis of the printed press. The surveys cover a wide range of topics, from the impact of social networks, the quality of the digital media, the willingness of citizens to pay for content and the degree of interactivity in the digital media for the future of journalism. Professionals and users agree that the news in 2020 will be better and more interactive than now. With regard to other matters, there is more disagreement. Professionals are much more critical than users when it comes to evaluating the quality of the news. The professionals’ survey was more qualitative, and, in their opinion, overcoming the crisis in the written press implies, among other measures, promoting analysis, prioritising quality over quantity and offering “niche” specialisation.

Laurent Vogel

In these uncertain times, saying that journalists are getting bad press would be an understatement. Political leaders, whatever their colour, are now starting to publicly pillory them. Press representatives used to be a homogeneous, urbane, trendy caste, pacing the corridors of power, frequenting the cocktail bars. For the last few years, the increased presence of star TV newscasters in glossy magazines and the omnipresence of certain columnists in talk-shows have obviously not helped boost the profession’s image among the public. At the side of this absolute minority of VIP journalists, the vast majority modestly pursue their work of informing the public, well away from the spotlights. Most of them still make a decent living out of it, though a growing minority is not managing to earn a decent wage. The phenomenon of long-term trainees and bogus self-employment is nothing new. For many years now, many budding journalists have seen this as the only way of getting a foot on the career ladder. After all, haven’t they got to earn their spurs? The problem these days is that such “probationary periods” are tending to become a permanent status. To such a point that budding journalists – and not just them – just can’t imagine gaining an open-ended employment contract. Looking specifically at the printed press, the Internet has become synonymous with free access to information, accelerating the decline in sales. The 2008 crisis triggered a collapse in advertising revenues, which newspaper owners and managing editors – more and more of them business school graduates – used as a pretence to “streamline” their operations. The consequence: newsrooms are empty, the “survivors” are on the brink of exhaustion, and much of a newspaper’s content is supplied by an army of invisible freelancers. In Europe, freelancers now constitute one-third of journalist union members. Isolated, having to compete against each other – and even sometimes against the hordes of non-professional unpaid “informers” churned out by the social media –, they are prepared to do everything to build up a reputation and work with certain media in the long term. Despite their extreme precariousness, they are little inclined to join forces. “To work in such professions, you have to pay: you pay for the enjoyment you get out it, you pay for the prestige they give you”, stated two (precarious) authors nearly ten years ago in a book highlighting the impoverishment of certain intellectual professions, including journalism (Les nouveaux intellos précaires, publ. Stock, 2009). Since then, the situation has got a lot worse, to the point that journalists are beginning to question the role and sense of their profession, with a growing percentage of them even thinking about doing something completely different. One cloud more gathering over our democracies … in these uncertain times.

Aukse Balcytiene

Javnost - The Public

Peter Laufer

RELATED PAPERS

Jurnal ABDINUS : Jurnal Pengabdian Nusantara

Syarif Hidayatullah

Stuart Turville

The Journal of the Korean Academy of Periodontology

Muhammad Sa'ad

Kathleen Newland

Archives of General Psychiatry

David Fleck

International Journal of Security and Its Applications

sushila madan

Andre Borba

在线办理concordia学位证书 康考迪亚大学毕业证学历学位证书学费发票原版一模一样

2015 IEEE Global Humanitarian Technology Conference (GHTC)

Bibek Gautam

Journal Of Oral Medicine And Dental Research

Ortega Pinto

Ruibin Geng

Maria Wurzinger

Neurosurgical focus

Shelly D . Timmons

Marcelo Thielo

Journal of Chemical Education

Malcolm Stewart

Food Science and Technology Research

Hideki Horie

Item: revista de biblioteconomia i documentació

Daniel Cassany

behnoush alimadadi

Guillermo Jamaica

RELATED TOPICS

•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

College of Agricultural, Consumer and Environmental Sciences

• From the Dean
• Overview of ACES
• International Programs
• Applying to ACES
• Financial Aid & Scholarships
• Future Students
• Program Explorer
• Undergraduate Degrees
• Graduate Degrees
• Online Programs
• Certificates
• Pre-College Programs
• Student Opportunities
• Study Abroad
• Current Students
• Career Services
• Research Areas
• Voices of ACES Blog
• ACES Newsletter
• Connect with ACES
• Alumni Association
• Alumni Events
• ACES Family Academies
• Alumni Awards
• Get Involved
• Alumni News
• Make a Gift
• Ways to Give
• 4-H Foundation

Visiting a country rich in love, faith and hope

Sierra Leone — nicknamed Salone by many who live there — is one of the poorest countries in the world. But it’s rich in other ways … with love, strong faith, and hope for the future. I spent my spring break in Freetown, the city where freed slaves returned to Africa.

With five Illinois College of Media journalism students and two professors, we shot b-roll, conducted interviews, and collected audio to create a documentary on mitigating period poverty.

Families in Salone constantly face impossible decisions. Financial constraints force households to choose between putting food on the table and purchasing period products.

With the culture of menstrual shame in the country, women are not supposed to speak about their cycle. It is seen as taboo and solely a female issue. Throughout the week, we spoke with sexual and reproductive health educators and advocates. These empowering women are spreading the message that periods are not something to be ashamed of but celebrated. There would be no human population without menstrual cycles.

Our team followed Uman Tok, an organization providing reproductive health awareness, education, and sustainable feminine health kits to women and girls. We also spoke with educators at the Education For All school; a nurse midwife who worked during the country’s civil war; and young Sierra Leonean women who have just reached puberty. It is my privilege to be able to share the stories of the most amazing, resilient, and kind-hearted women I have ever met.

The documentary was made in partnership with journalism students at Fourah Bay College at the University of Sierra Leone. Working with students from Freetown, who are my age and also studying journalism, was an experience like no other. The close relationships, formed only after a week, make me long to return to Salone as soon as possible.

Until then, our team of Illinois journalism students will do all we can to advocate for our peers in Africa.

The Sierra Leone study abroad team presented a trailer of the documentary on April 25, 2024, at the Undergraduate Research Symposium at the Illini Union. The documentary will air today, May 28, 2024, Menstrual Hygiene Day.

Elena Cleary is a student in Agricultural Leadership, Education and Communications.

Read more from the College of Media

Advertisement

After Trump’s Conviction, a Wary World Waits for the Fallout

Already braced for uncertainty about the U.S. election, countries in Europe and Asia are now even more unclear about the future of American diplomacy.

• Share full article

By Hannah Beech and Paul Sonne

• May 31, 2024

The world does not vote in American presidential elections. Nor do its jurors play a part in the American judicial system. Nevertheless, the conviction of Donald J. Trump on all 34 felony counts in a hush-money trial in a New York court on Thursday has again made clear how consequential what happens in the United States is for the rest of the planet.

Many America-watchers are grappling with the same questions posed by people in the United States: Can Mr. Trump still run for president? (Yes.) And if so, will the guilty verdicts cut into the support from his political base? (Unclear.)

Foreign observers also began wondering if Mr. Trump, already a volatile force, would become even less likely to stay within the guardrails of normal politics and diplomacy if he won the presidency again in November.

Mr. Trump’s supporters in anti-immigrant, right-wing nationalist circles abroad quickly jumped to his defense. Viktor Orban, Hungary’s Kremlin-friendly prime minister, called Mr. Trump “a man of honor” in a post on X and said the American people should deliver their own verdict in November.

Matteo Salvini, Italy’s deputy prime minister and the leader of the hard-right League party, expressed “solidarity and full support,” and called Mr. Trump a “victim of judicial harassment.”

“This verdict is a disgrace,” Nigel Farage, the pro-Brexit campaigner and Trump supporter, who is honorary president of Reform UK, a small right-wing party in Britain, wrote on social media. “Trump will now win big.”

President Vladimir V. Putin of Russia did not immediately respond to the verdict but has seized on the situation more broadly to undermine American influence. Mr. Putin last year called the various proceedings against Mr. Trump political “ persecution ” and said they had revealed the “rottenness of the American political system, which cannot pretend to teach others about democracy.”

His spokesman, Dmitri S. Peskov, reiterated the point on Friday in response to the verdict, saying it was clear to the entire world that the U.S. authorities were trying to eliminate political rivals “by all possible legal and illegal means.”

The convictions by a Manhattan jury come as the question of American engagement has become central in several global crises.

In Ukraine, the war effort against Russia has been stymied after Republicans in Congress delayed American military aid for months.

In Europe, leaders reliant on the United States for their defense are jittery about a return to a more acrimonious relationship with Washington and a possible withdrawal of American support for hardening defenses against Russia.

In Asia, where the Biden administration perceives a growing Chinese threat and worries about a possible invasion of Taiwan, American allies are concerned about the sanctity of defense treaties that have long girded the regional security order.

On the campaign trail, Mr. Trump has said he would encourage Russia to attack any NATO member that doesn’t pay sufficiently for its defense and has questioned whether the United States should defend South Korea, a treaty ally that hosts a large American military presence. He is considering the Ohio senator J.D. Vance, one of Washington’s most vociferous opponents of military aid for Ukraine, as a possible running mate.

Foreign analysts worry that Mr. Trump’s favored currency, unpredictability, could again shake up the global order.

Concern about his possible return to the White House is particularly palpable in Germany, the object of Mr. Trump’s ire for much of his first term and the host of more than 35,000 U.S. troops.

Andrea Römmele, vice president of the Hertie School, a public policy-focused graduate school in Berlin, said many Germans watching the Trump verdict were relieved to see that even a former president was not above the law in the United States. But she said Germans remained very anxious about a Trump victory.

“I think everyone is much more prepared to think the unthinkable,” she said.

Prime Minister Donald Tusk of Poland, whose right-wing domestic opponents accuse him of using the judiciary to settle political scores, hailed the conviction of Mr. Trump in New York as “an American lesson” for Polish politicians.

“The law determines guilt and punishment, regardless of whether the perpetrator is a president or a minister,” Mr. Tusk said in a message posted on X. A veteran centrist, Mr. Tusk took office after an October election that ousted a nationalist government that cultivated close ties with Mr. Trump during and after his time in the White House.

Still, on Friday, most foreign governments, forced to surf every shift in the American political mood, reacted cautiously.

“I would like to refrain from commenting on matters related to judicial procedures in other countries,” Yoshimasa Hayashi, Japan’s chief cabinet secretary, said at a news conference in Tokyo on Friday.

In Britain, where a national election campaign is underway, Prime Minister Rishi Sunak refused to discuss the Trump case. His Labour Party opponent, Keir Starmer, a former top prosecutor, said he respected the court’s decision and called the situation unprecedented.

“Ultimately whether he is elected president will be a matter for the American people and obviously, if we’re privileged to come in to serve, we would work with whoever they choose as their president,” Mr. Starmer told BBC Radio Scotland.

Mao Ning, a spokeswoman for China’s foreign ministry, declined to comment on the verdict. She said she hoped whoever was elected president would “be committed to developing healthy and stable China-U. S. relations.”

The possibility of Mr. Trump’s return to the White House is a source of anxiety for U.S. allies in Asia that rely on Washington for their defense.

When Prime Minister Fumio Kishida of Japan made a state visit to Washington in April, President Biden called relations between the countries the most important bilateral alliance in the world. With American concern rising over China’s expanding military footprint, Mr. Biden has strengthened American defense partnerships with Japan, South Korea, the Philippines and others in Asia.

By contrast, while president, Mr. Trump called for Japan, which hosts more than 50,000 American troops on its soil, to pay $8 billion for the upkeep of American bases there. (It never happened.)

Still, the fundamental tension in regional geopolitics — the contest between the United States and China — will continue no matter who wins the American presidential election.

“Beijing has no illusion about Trump or Biden, given their anti-China solid stance,” said Lau Siu-kai, an adviser to the Chinese government on Hong Kong policy. “Beijing is all set for a more intense confrontation with the U.S. over technology, trade and Taiwan.”

Officials in China’s embassy in the United States and its consulates around the country are most likely scrambling to assess how the verdict could affect the election, said Willy Lam, an analyst of Chinese politics at the Jamestown Foundation in Washington.

“The majority of Xi Jinping’s advisers now think a Trump presidency might be worse for U.S.-China relations,” Mr. Lam said of China’s top leader. “If Trump were to win, given the now peculiar circumstances of his victory, he might gravitate towards unpredictable actions to assert his authority.”

There is a sense in Asia that the region is perennially overlooked and underappreciated by U.S. presidents, particularly as crises in Europe and the Middle East have monopolized Mr. Biden’s attention. That sentiment was also felt acutely during Mr. Trump’s presidency, and for American partners in Asia it was made worse by his affinity for regional strongmen.

In addition to occasional expressions of admiration for Mr. Putin and Kim Jong-un of North Korea, Mr. Trump invited to the White House a former army chief who led a coup in Thailand and installed himself as prime minister. Mr. Trump drew accolades from Rodrigo Duterte, formerly the president of the Philippines and now under investigation by the International Criminal Court over his deadly war on drugs.

The Philippines is now led by the son of the longtime dictator Ferdinand E. Marcos, who died in exile in Hawaii. He has reoriented the country away from China back toward the United States.

In at least one regard — the prosecution of former leaders — the rest of the world is far ahead of the United States. South Korea, where four former presidents have been convicted of corruption and abuse of power, has made something of a national sport of imprisoning disgraced leaders. The former French presidents Nicolas Sarkozy and Jacques Chirac were convicted of corruption.

Jacob Zuma, the former president of South Africa, has been charged with money laundering, among other crimes. And Luiz Inácio Lula da Silva was sentenced to years in prison for corruption after leading Brazil. His convictions were eventually annulled. He is again president of the country.

Reporting was contributed by Stephen Castle, Elisabetta Povoledo, Roger Cohen, Zixu Wang, Andrew Higgins, Camille Elemia , Choe Sang-Hun , Motoko Rich , Alexandra Stevenson , Sui-Lee Wee and Sameer Yasir .

An earlier version of this article misstated the length of Rodrigo Duterte’s term in office. It was six years, not eight years.

How we handle corrections

Hannah Beech is a Times reporter based in Bangkok who has been covering Asia for more than 25 years. She focuses on in-depth and investigative stories. More about Hannah Beech

Paul Sonne is an international correspondent, focusing on Russia and the varied impacts of President Vladimir V. Putin’s domestic and foreign policies, with a focus on the war against Ukraine. More about Paul Sonne

Our Coverage of the Trump Hush-Money Trial

Guilty Verdict : Donald Trump was convicted on all 34 counts  of falsifying records to cover up a sex scandal that threatened his bid for the White House in 2016, making him the first American president to be declared a felon .

What Happens Next: Trump’s sentencing hearing on July 11 will trigger a long and winding appeals process , though he has few ways to overturn the decision .

Reactions: Trump’s conviction reverberated quickly across the country  and around the world . Here’s what voters , New Yorkers , Republicans , Trump supporters  and President Biden  had to say.

The Presidential Race : The political fallout of Trump’s conviction is far from certain , but the verdict will test America’s traditions, legal institutions and ability to hold an election under historic partisan tension .

Making the Case: Over six weeks and the testimony of 20 witnesses, the Manhattan district attorney’s office wove a sprawling story  of election interference and falsified business records.

Legal Luck Runs Out: The four criminal cases that threatened Trump’s freedom had been stumbling along, pleasing his advisers. Then his good fortune expired .