Felix M. Simon

Generative AI and news report 2025: How people think about AI’s role in journalism and society Our survey explores how people use generative AI in their everyday lives, what they think its impact will be on different areas of society, and what they think about its use in news and journalism spe...

If you are interested in how this all varies by country (in some cases, it does), check out the @reutersinstitute.bsky.social report, by @felixsimon.bsky.social @rasmuskleis.bsky.social and myself.

reutersinstitute.politics.ox.ac.uk/generative-a...

Finally, some data on trust. It's difficult to interpret data on trust in gen AI systems at the moment because there is limited public awareness - with the exception of ChatGPT, 50% or more have not even heard of them.

No surprises that there's an age gap in the use of gen AI, with use more widespread among younger people.

But this only applies to standalone systems like ChatGPT. For AI embedded in other products, like Meta AI and Copilot, there's no age gap because the host product is used by people of all ages.

On the use of specific systems, ChatGPT is still the most widely used (22% weekly) - ahead of Gemini (11%), Meta AI (9%) and Copilot (6%).

Worth remembering that the public's use of a lot of the tools favoured by experts, like Claude and Perplexity, is still very marginal - 1% weekly.

A thread on how people's use of generative AI has changed in the last year - based on survey data from 6 countries (🇬🇧🇺🇸🇫🇷🇩🇰🇯🇵🇦🇷 ).

First, gen AI use has grown rapidly.

Most people have tried out gen AI at least once (61%), and 34% now use it on a weekly basis - roughly doubling from 18% a year ago.

✨🤖 Check out our new research on AI use around news and information and attitudes towards AI in society and journalism – with @rasmuskleis.bsky.social & @richardfletcher.bsky.social

Finally, as search engines increasingly integrate AI generated answers, we asked about trust in these – the trust scores are high across the board, with higher net positives than any of the standalone tools.

All this and more in the report here reutersinstitute.politics.ox.ac.uk/generative-a... 3/3

Asked whether they trust different AI tools, 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 2/3

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

More optimists than pessimists for e.g. science and search engines, but more pessimists than optimists for news media, government, and – especially – politicians 1/3

And ppl can still think something is "good" or "good enough", even if they don't fully trust it I guess. But we barely scratched the surface here, lots of follow up questions to be asked.

As for your point on use: Well, the other features still work so seeing an AI-generated overview that you don't think is great will not deter you from using the rest of Google Search (apart from the fact that it's baked into so much infrastructure, so convenience argument applies, too).

Thanks so much, Andy! I think seeing relatives use it (but also seeing myself use it) helped anchor my expectations here. "We (you and me)" are less trusting because we study this stuff all day long. And these answers can be correct and ppl can see that, so good reason to trust them some of the time

Thanks, Hannes!

Thank you, David :-) @richardfletcher.bsky.social will be be able to confirm, but I'm afraid the answer is "No" given the way we asked. Given the primacy of the big systems, one could reasonably speculate it will mostly be one of them, however.

(And yes, read the report ;-)

AI is your newest audience: The B2A(2C) design challenge B2A2C — Business to Agent to Consumer — is the new opportunity for information design. AI is reshaping the media pipeline and translation layers are the new source of power. Here’s what media makers ...

@shuwei.bsky.social wrote something interesting on this recently splicemedia.com/stories/ai-i...

I am also thankful to Masaharu Ban, Gretel Kahn, Priscille Biehlmann, Magnus Bredsdorff, and Tania Montalvo for their advice on the translations & @mitalilive.bsky.social, Kate Hanneford-Smith, Alex Reid, @eduardosuarez.bsky.social, Rebecca Edwards for helping to move this project forward.

@michelledisser.bsky.social helped with red-teaming the questionnaire and the interpretation of some of the results.

A big thank you to my co-authors @richardfletcher.bsky.social & @rasmuskleis.bsky.social for all their hard work. Caryhs Innes, Xhoana Beqiri, and the team at @yougov.co.uk were invaluable in fielding the survey and the @reutersinstitute.bsky.social research team helped with approaching the topic.

So much for now. We will be diving deeper into this across our newsletter and I will try and highlight some specific areas. All of the above and more can be found in the full 60 page report.

Here is the link again: reutersinstitute.politics.ox.ac.uk/generative-a...

Likewise, and perhaps again not popular new for many outlets, seeing AI labelling on news is infrequent relative to daily news use. Only 19% see AI labels daily and 28% weekly – a low number considering that 77% say they use news daily.

And despite a growing number of outlets introducing audience-facing AI, most people don’t yet recall seeing these AI features. Sixty per cent say they do not regularly see AI features on news sites or apps. Most common are AI summaries (19%) and AI chatbots (16%).

People have limited confidence in human oversight of AI in news. Only 33% think journalists ‘always’ or ‘often’ check AI outputs before publishing. Trust in news strongly correlates: 57% of those who ‘strongly trust’ news think such checks happen (19% strong distrusters).

… but there is also the view that AI will make news less transparent (−8) and less trustworthy (−19); and these views have hardened since 2024 (we saw no decreases).

People continue to have mixed expectations about what AI will do to news. As with last year, many assume AI will make news cheaper to produce (+39 percentage point difference between those that said more and those that said less) and more up to date (+22)…

Mind you, this is self-reported and people might not always know that AI has been in used in something they are consuming, but it shows that the public is sceptical of the use in news (but as the report shows, it depends on the context).

Finally, looking at AI in the news, there is a clear ‘comfort gap’ between AI- and human-led news. On average, only 12% say they are comfortable with news made entirely by AI; (21% with a ‘human in the loop’, 43% when a human leads, 62% for entirely human-made news).

Only three sectors see the pessimists outnumber the optimists when it comes to AI and its use by these actors – news media, government, and, especially, politicians and political parties…

On AI in society findings, Rasmus will have more to say on this but generally, there are more optimists than pessimists when it comes to AI – especially for AI use in sectors like healthcare, science, and search engines (!).

Among those who have encountered AI answers, 50% say they trust them. Respondents emphasised their speed and convenience and the fact that AI aggregates vast amounts of information as reasons to trust them, although trust seems conditional.

Self-reported click-through behaviour is mixed. Among those who saw AI answers, 33% say they always/often click links in AI overviews, 37% say they do so sometimes, and 28% rarely or never click through. Younger people & those who trust AI search are more likely to say they do