Generative AI news discovery looks much more concentrated than other forms of access

Posted on May 22, 2026 | Leave a comment

Generative AI could in principle feature a great diversity of sources in output responding to news-related queries.

But do various generative AI products actually do this?

In early 2026, Roa Powell and Carsten Jung published a piece of research for the Institute for Public Policy suggesting maybe not – based on analysis of responses to a sample of 100 hypothetical UK news queries submitted to four different AI tools (ChatGPT, Perplexity, Gemini, and Google AI overview), they found that these tools draw “on a narrow range of prominent news brands”.

As a benchmark for interpreting their results, I have created the slide below.

The figures in red are my calculations of the Herfindahl–Hirschman Index (HHI) (a common measure of market concentration) for each of the four AI tools based on the share of reference that goes to each of the top ten news brands (as reported by Powell and Jung).

The figures in black are taken from this piece of work I did with Richard Fletcher included for comparison, namely similar HHI calculations based on historic data from the UK from 2017 for online news accessed directly, via search, news aggregators, or various kinds of social media, as well as, for further comparison, HHI figures for television viewing and weekly print newspaper circulation at the time.

The concentration of attention, with a few brands accounting for a very large share of reference, is far greater for the AI tools than any other form of access.

While the underlying data is not like-for-like comparison (the bulk of the 2017 data is based on passive tracking of actual UK users, the 2026 AI tool data is generated by prompting), I think the figures are still interesting and thought-provoking. (And not necessarily unique to the UK – Nikos Smyrnaios and Olivier Koch has published a piece of work, based on a somewhat different methodology, suggesting results for France that are also about the 2,000 bar for a highly concentrated market.)

It’s not just that Powell and Jung are right to stress that a narrow range of prominent news brands (some of whom have commercial deals with the AI companies) loom very large in AI output.

It is also that the concentration in question, measured here in terms of share of reference, is far, far greater than any of the different kinds of access we analyzed based on the 2017 data – even more concentrated than direct access, which itself was significantly more concentrated than any other kind of access at the time.

We already know that AI tools generate far fewer referrals than established platforms do – worrying enough for a multitude of publishers competing to connect with the public.

It also seems the (comparatively fewer) referrals they do drive are highly concentrated amongst a select few publishers, much more so than has been the case for search, social, or news aggregators.

(Underneath the topline, there is some diversity from tool to tool (just as we found different outlets doing well via different platforms), as Powell and Jung writes, their findings suggest “each AI tool prioritizing news brands in different ways, in each case foregrounding a distinct selection of news outlets compared with those that are currently most popular across the UK” – go read their report for more details.)

Leave a comment

Posted in Business of journalism, digital politics, Trivia and impressionism

Tagged , , , ,

Joining Council of Europe Expert Committee on media regulation and platforms

Posted on February 10, 2026 | Leave a comment

I am looking forward to beginning work in the Council of Europe’s new Expert Committee on Media Regulators in a Platform-Based Environment. There are three reasons I think this is important.

First, with the other independent expert members, I hope we can ensure that we can help base the committee’s work on research. Policy is necessarily political, but it is also about evidence, and here independent experts can help.

Second, media regulators are typically set up for a mass media environment, but today have to do their important work in a platform-dominated environment. I hope we in the committee can offer ideas for how to evolve regulators’ role, remit, and resources to the media environment we have, rather than the one we had.

Third, the Expert Committee’s work is under the aegis of the CoE Division in charge for work on freedom of expression and media freedom. This is particularly important in a time of democratic backsliding and attacks on fundamental rights, as media regulators are unfortunately often one of the independent institutions that would-be autocrats seek to subvert and instrumentalize in pursuit of media capture and worse.

I hope our work in the committee can consider both the different kinds of good that media regulators can do, help ensure they have tools for the present and the future, not just the past, and consider how to limit the harm they can be used to do.



Looking forward to working with Rowena Burke Audun Aagre Stephanie Comey Tanja Kerševan Michèle Ledger Krisztina Rozgonyi Maria Luisa Stasi and everyone else involved in this over the next two years.

Leave a comment

Posted in Media policy

Tagged

Media subsidies for citizens – report out

Posted on October 29, 2025 | Leave a comment

’Mediestøtte for borgerne – demokratisering, fremtidssikring, og forenkling’ (Media subsidies for citizens – democratization, futureproofing, and simplification). The recommendations from the Commission on the Future of Media Subsidies that I have chaired for the Danish government is now out.

In our report, we have worked to develop a set of principles to undergird and practical solutions to deliver direct subsidies to news media in Denmark in a way that will modernize the current system and ensure it can work in an effective, legitimate, and transparent manner going forward.

For those interested in having a look at the full report in Danish, it is available here.

And stay tuned for more from me in English in the coming weeks, summarizing key parts of the approach we outline, as I hope our work will be useful elsewhere

It was a real privilege to chair the commission over the last year, and to work with the members and the civil servants who supported us.

Leave a comment

Posted in Media policy

Tagged

3 key findings from new report on generative AI use

Posted on October 7, 2025 | Leave a comment

How do people think different sectors’ use of generative AI will change their experience of interacting with them?

That’s one of the question we fielded in a new survey, and one of the three key findings from my perspective – looking across the six countries covered, there are more optimists than pessimists for e.g. science and healthcare, and for search engines and social media, but more pessimists than optimists for news media, the national government, and – especially – politicians and political parties.

Elsewhere in the survey, we ask whether people trust different generative AI offers – the picture is very differentiated, with net positive trust scores for e.g. ChatGPT, Google Gemini, and Microsoft Copilot, but negatives for those that are seen as part of various social media companies.

Finally, as search engines increasingly integrate AI generated answers, and more and more of us see these all the time, we asked about trust in these answers – the trust scores are high across the board, with higher net positives than any of the standalone tools. (With this and the question above, surveys do not measure whether the entities in question are trustworthy, and do not tell us anything about whether people should trust them, they provide data on whether they do trust them.)

Beyond that, the report, which I wrote with Felix Simon and Richard Fletcher, and which is published by the Reuters Institute for the Study of Journalism, is chock full of fresh data on generative AI use (basically doubled since last year), what people use these tools for (increasingly for information, presenting a very clear direct competition to search engines), and what they don’t (yet) use them for all that much (getting the latest news).

Leave a comment

Posted in Comparative media research

Tagged

US, big, or commercial? What do you want platform alternatives to?

Posted on June 23, 2025 | Leave a comment

I was asked about alternatives to dependence on dominant US American for-profit platform companies at an Internet Governance Forum today. Below my response – for those interested in more from me on the topic, I spoke at somewhat greater length about it at the Nordic AI in Media Summit back in April. More broadly, everyone interested in this topic should read the Eurostack pitch paper.

Below my response, video here.


There are plenty of options, but I think the real question is, we need to hold people in positions of power, including public and political power, to account in terms of how they understand the issue and whether they act accordingly.

The question here – when looking at dependence on US-American big commercial platform companies – is which part of that phrase you stress.

If you think the problem is that they’re US-American, then the path you pursue is obvious.

It is that you try to create national, or in the case of Europe, regional champions. And then when they have the right passport, and you are reliant on “grande technologie” rather than big tech, things are fine, right? Because then those companies are beholden to a different set of politicians. And then let’s just hope that whoever is the next inhabitant of the Élysée is not going to abuse that power the way that we see in some other cases. The question then question is whether we as [citizens] can expect very different behavior from large corporations who hold different passports. […]

Then the second way to think about the problem is that they are big.

Now then the alternatives are also, I think, quite clear. You’re thinking about decentralized, federated, open source solutions.

Now I think it needs to be very clear that very few people in positions of power seem to think this is the problem, because if they did, they would pursue those alternatives already, because they exist, like the Fediverse including Mastodon or LibreOffice. There are options in this space and we have now 25 years of revealed preference from people in positions of power. This is not what they want. So those alternatives exist, but they are not being pursued.

Then finally, of course, your analysis might be that the problem [with incumbent dominant platforms] is that they are commercial and that’s where we can turn to the possibility of public service alternatives.

And I think it’s possible to do this. It’s not easy. We need to decide what are they going to do? There are many layers of the stack one could look at. How are they going to be funded? This is not going to be cheap. Who’s going to make the rules and who’s going to enforce them? Like all the controversies we see around content moderation decisions. Imagine those only with the politicians in your country of origin making the decisions rather than Mark Zuckerberg and his Oversight Board.

The question then is a question of priorities, right? In Europe alone, we spend an estimated 40 billion euros a year on public service media. That has been stagnant, in some cases declining in recent years, but we could make investments of a similar size. Europe is a 20 trillion US dollar economy. Public spending in Europe alone is about 10 trillion euros a year. It’s a question of priorities.

And that’s why I think we really need to be clear about.

The full panel is available from the IGF on YouTube and was a great discussion with really interesting participants.

Speakers:

  • Kjersti Løken Stavrum, Chairman of the Board, CEO, Schibsted, Tinius
  • Anine Kierulf, Associate Professor, UiO and the Norwegian National Human Rights Institute
  • Rasmus Kleis Nielsen, Professor, Uni. Copenhagen, Reuters Inst. for the Study of Journalism
  • Chris Disspain, Former Vice-Chair of the Board of ICANN, Chairman of DNS Capital Ltd, author, lawyer   
  • Anya Schiffrin, Director, Tech., Media and Comm., Columbia University
  • Tawfik Jelassi, Ass. Dir.-General/Deputy Dir.-General, UNESCO
  • Pamella Sittoni, Executive editor/Managing Editor, Daily Nation/Nation Media Group (prev)

Moderator:

  • Helle Sjøvaag, Prof. & Vice Dean, University of Stavanger

Leave a comment

Posted in Conferences

Tagged ,

2025 Digital News Report out

Posted on June 19, 2025 | Leave a comment

“Alternative media voices often have a wide reach and appeal to audiences that news publishers have been keen to engage with but the report also shows that, when it comes to underlying sources of false or misleading information, online influencers and personalities are seen as the biggest threat worldwide along with national politicians”, Mitali Mukherjee writes in her introduction to the 2025 Digital News Report.

There are countries where news media still have a strong connection with much of the public, and where publishers have adapted well to a challenging digital media environment (including my native Denmark), but overall the report is a sobering read for the news media. As lead author Nic Newman writes: “In most countries we find traditional news media struggling to connect with much of the public, with declining engagement, low trust, and stagnating digital subscriptions.”

In the report, we document how platforms are increasingly central to how many find and access all sorts of content, including news, as well as a continued fragmentation of the platform space. There are now six networks with weekly news reach of 10% or more compared with just two a decade ago. Instagram, WhatApp, and TikTok in particular have grown in importance, whereas BlueSky still only has tiny reach amount our respondents.

While industry data suggests X is much diminished in terms of how intensely it is used, survey data on weekly use – perhaps surprisingly – suggests stable reach overall. A liberal exodus seems to have been matched by a growing number of right-wing users, and after many years of having a predominantly left-wing user base, X now has slightly more right-wing users.

In this increasingly distributed and platform-dominated environment, large parts of the public continue to be concerned about what is real and what is fake when it comes to online news – when asked what they are most concerned about, domestic politicians and online influencers/personalities top the list, in terms of platforms, concern is focused on Facebook and TikTok.

First Elon Musk and later Mark Zuckerberg has said they want to reduce how much content is subject to moderation on their platforms – while some political actors may applaud this, it is not clear the public does. A plurality in many countries say they want more harmful or offensive content removed from social media.

Finally, as generative AI is increasingly widely used, integrated into platforms, and adopted by many news publishers, we asked respondents what they think this will mean for news content – while there is some optimism AI-powered news will be more up to date and easier to understand, the topline is people expect it to be cheaper to make but less trustworthy.

All that and more in the 2025 Digital News Report, with topical chapters, country pages, interactive data, and more on the Reuters Institute website – an incredible team effort that I am proud to be part of.

Leave a comment

Posted in Comparative media research

Tagged , , , ,

What could ‘European alternatives’ mean? – NAMS keynote on possible platforms

Posted on April 29, 2025 | Leave a comment

I gave the closing keynote at the 2025 Nordic AI in Media Summit April 24 under the title “What could ‘European alternatives’ mean? Three possible platform models in search of your support”.

Photo credit: Philip Jørgensen

As a scientist, I prefer to deal in reliably, empirical knowledge, but I was grateful to be invited to think aloud in terms of possible responses to the current moment in geopolitics and tech.

If you are interested in my three possible models, each of which represent an approach trying to solve for a different definition of the problem ( (1) reliance on American tech? Airbus-for-the-internet as national/European champions!, (2) reliance on for-profit tech companies? BDC/public service platforms! (3) reliance on big technology companies? Mastodon as a decentralized, open-source alternative!).

In each case, for every alternative, at every level of the stack, at least three questions need clear answers – what, exactly, is the alternative meant to do, who will fund it, and how is it going to be governed.

It all strikes me as a wicked problem akin to climate, defense, and the future of strained welfare systems – and of a comparable scale and scope, and seriousness that requires serious responses. I hope the models I outline and the questions I offer can help structure how we discuss possible responses and move beyond the declarations and rhetoric that suffice for headlines and a bit of publicity, but don’t actually change anything.

Video of my talk below.

Leave a comment

Posted in Conferences, Talks and presentations

Tagged ,

“Avoiding the news” turns one

Posted on December 20, 2024 | Leave a comment

Wrapping up the year grateful for how people have engaged with “Avoiding the News”, the book Ben Toff, Ruthie Palmer, and I published at the beginning of the year. Sharing it with the world has, as we hoped it would be, been the beginning of a wider conversation, not the final word.

In the book, we draw on hundreds of interviews as well as survey data to show that news avoidance is not “just” a response to the content on offer in the news. but also fundamentally shaped by who we are, what we believe, and the tools we rely on. It happens at the intersection between identity, ideology, and infrastructures, and tends to compound existing inequalities.

Ruthie Palmer and I on stage at the World News Media Congress, discussing the book with Amalie Kestler and Shazia Majid.

We have had numerous conversations with journalists and editors concerned about the fraying connection between much of the news and much of the public, a chance to discuss the book with participants at the International Journalism Festival, with experienced journalists and editors at the World News Media Congress, and seen reporters from the New York Times, the Washington Post, and elsewhere engage with the work. There also continues to be a vibrant community of academics researching the phenomenon in various ways, some of them contributing to a special issue of Journalism Studies on the topic.

Ben Toff presenting during our panel session at the International Journalism Festival.

The book has also been named a Choice Outstanding Academic Title, an award given by the American Library Association’s Choice Reviews to “outstanding works for their excellence in presentation and scholarship, the significance of their contribution to the field, their originality and value as an essential treatment of their subject”.

We hope journalists and those who care about journalism will continue to engage with the analysis we present and the issues we identify.

If you are curious about the book, you can read an excerpt here where we identify the groups more likely to be consistent news avoiders, if you are interested in our thoughts on how journalists could respond, we discuss some options here.

The book is available here from the publisher, Columbia University Press, with a 20% discount using the code CUP20SM.

Leave a comment

Posted in Uncategorized

Three highlights from Ditchley discussions

Posted on December 9, 2024 | Leave a comment

Three highlights from a weekend spend discussing “The Role of the Fourth Estate in Democracies” at the Ditchley Foundation.

* The problem is not the public – people are interested in the world around them, they use media to explore it, make sense of it, and navigate it, and engage with content and information they find useful and relevant.

* Transitions are hard for news media, but also demonstrably possible – leaders from a range of very, very different kinds of publishers were very clear they think believe much of the current industry is at best in managed decline, but also that they see a range of sustainable paths forward for the business of news.

* Part of that is the useful discipline of having to reach your audience – several participants spoke in different ways about how the shift from direct discovery through channels dominated by news publishers to distributed discovery through platforms while in many ways very challenging has also forced a new, healthy, humility on journalism and the news media by making clear they can never take the public and people’s attention for granted.

All contributions are confidential but under the Ditchley rules we can draw on the substance of it as long as we do not disclose who said what – I was glad that this year’s discussions were much more respectful of the public and more cautiously optimistic about the future of (parts of) the news industry than many such discussions I have been part of over the years.

Leave a comment

Posted in Uncategorized

New project: “Power over Platforms?”

Posted on November 11, 2024 | Leave a comment

I am very grateful that the Danish National Science Foundation has awarded me a DNRF Chair grant to support my new project “Power over Platforms?”.

Power over platforms?

The aim of this 3-year project is to understand how power is exercised over platform companies such as Google, Meta, and their competitors by actors who have neither raw economic nor formal regulatory or legislative means.

The focus is on how civil society groups, interest groups, professional associations, and companies from other sectors sometimes actively try to shape how platform companies operate. The starting hypotheses are that these actors (a) do this because they believe they are able to exercise at least some influence and (b) this is sometimes the case.

Systematically analyzing the actors involved in trying to influence key decisions made by different platforms on key issues—content moderation, privacy, and the use of generative AI for political information and speech—across different jurisdictions (the US and UK as major markets outside the EU, key markets inside the EU), and across different platforms (primarily consumer-facing content platforms including Google, Meta, and their smaller competitors TikTok, X, Snapchat, and Reddit) the project sets out to identify who seeks to influence platform governance, how, and what the outcomes are across countries and across companies.

The project builds on and goes beyond previous work on the “Power of Platforms” I did with Sarah Anne Ganter, and seeks to expand our scientific understanding of platform governance by analyzing a wide range of actors involved, some of whom have received limited attention from researchers. It aims to provide insights which can in turn help inform public and policy discussions about how to respond to the role that platform companies play in a range of important areas including free speech (through content moderation), privacy (data protection and encryption), and, with rapid development and deployment of generative artificial intelligence, new ways for people to exercise their fundamental right to receive and impart information and ideas.

This is an important area to research because platform companies increasingly develop and enforce principles, policies, and practices that go above and beyond what is legally required in defining what they consider acceptable behavior and content. But who, in turn, seeks to and sometimes manages to influence how the companies do this? When do they succeed? These questions are at the heart of the project.

I will be working with two postdoctoral researchers on the project (interested in these posts? Details here – apply by December 13) and various international collaborators.

I am grateful to the Danish National Research Foundation for deciding to fund the project, and to everyone in the Department of Communication and at the University of Copenhagen more broadly who have supported me and helped with the application process.

Leave a comment

Posted in Uncategorized

Tagged , , , , ,