More people get news from AI chatbots, but trust remains low

Weekly use of AI chatbots for news has risen from 7% to 10% globally, according to the Reuters Institute's Digital News Report 2026, but trust remains low at 20% among the general population. Only 4% of users regularly click through to original sources, raising concerns about the impact on journalism and public discourse.

More people get news from AI chatbots, but trust remains low Key Points - Global use of AI chatbots for news has risen from 7 to 10 percent, according to the Reuters Institute's Digital News Report 2026. - While 44 percent of active users trust AI-generated news, only 4 percent regularly click through to original sources. - Chatbots risk reinforcing users' existing beliefs and splintering public discourse through hyper-personalized content, but they can also make complex topics more accessible and expose people to a wider range of perspectives. The Reuters Institute's Digital News Report 2026 finds that weekly use of AI chatbots for news has climbed from 7 to 10 percent globally. AI tools like ChatGPT and Google Gemini are playing a bigger, though still small, role in how people get their news. That's the takeaway from the Reuters Institute's Digital News Report 2026. Weekly chatbot use for news rose from 7 to 10 percent worldwide. Just 1 percent of respondents call AI chatbots their main news source. The growth is mostly coming from markets in Asia, Africa, Latin America, and Southern and Eastern Europe. Young, news-hungry users are leading the charge Chatbot news use skews young and engaged, according to the study. Among 18- to 24-year-olds, 17 percent use chatbots for news, compared to just 5 percent in the oldest age group. The 25-to-34 bracket saw the strongest relative growth, jumping 4 percentage points. Usage among self-described "news lovers" hits 18 percent, well above the 7 percent among casual consumers. People with extreme political views also use AI chatbots for news more often: 16 percent on the far left and 15 percent on the far right. Researcher Dr. Amy Ross Arguedas says these groups simply tend to be more interested in news. Across the 45 markets surveyed, asking follow-up questions is the top use case at 42 percent. That's followed by getting current news 35 percent , summaries 34 percent , checking the reliability of news sources 33 percent , and simplifying news 30 percent . In markets with low press freedom scores, like Hong Kong and Turkey, and in markets where trust in news is low, like Hungary and Romania, using chatbots to check source reliability ranks especially high. Globally, 42 percent of users say they want more depth or explanation. Another 39 percent say AI is faster than other ways of getting news. More use, more trust, but the baseline is still low Only 37 percent of respondents trust most news. Trust in news from AI chatbots sits at just 20 percent among the general population. But the picture shifts among actual users: 44 percent of chatbot users trust AI-generated news, compared to only 17 percent of non-users, according to the report https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026/emerging-uses-ai-chatbots-news-and-what-it-means-journalism . At the market level, the data shows a strong link between trust and usage; much stronger than on social media. The study attributes this to the fact that using chatbots for news is a more deliberate choice than the often passive way people stumble across news on social media. Almost nobody clicks through to original sources Across all respondents in 27 markets, only 4 percent say they always or often click from AI chatbots to original sources. For search engines, that number is 19 percent. For social media, 17 percent. The gap partly reflects chatbots' much smaller user base for news, and, of course, how the system works. If I get an answer insteado of an link, of course I have less incentive to click and this shows in the data. This confirms earlier studies https://the-decoder.com/pew-finds-that-only-1-percent-of-users-click-a-source-link-directly-from-googles-ai-overviews/ and puts Google's defense of its AI Overviews into perspective. In a legal dispute https://the-decoder.com/landmark-german-ruling-declares-googles-ai-overviews-are-googles-own-words-and-makes-it-liable-for-false-answers/ , the company argues that people can simply verify the source of false AI-generated claims, just like they do with traditional search. They could. But they don't. And the cited sources don't always match the answer anyway. When chatbot users do click, they're less likely to seek more detail 51 percent than search engine users 59 percent or social media users 60 percent . Instead, they're more likely to click to verify information or learn more about the source. The study's recommendation for publishers: don't try to compete with AI platforms on their own turf. Focus on what chatbots can't deliver - original reporting and journalistic credibility. Confirmation bias and a splintering public sphere Chatbots can misrepresent source material. That's a well-known risk https://the-decoder.com/leading-ai-chatbots-are-now-twice-as-likely-to-spread-false-information-as-last-year-study-finds/ . But two other dangers may be bigger. The first is sycophancy https://the-decoder.com/sycophantic-ai-chatbots-can-break-even-ideal-rational-thinkers-researchers-formally-prove/ . AI chatbots tend to confirm what users already believe rather than push back. Ask a news question with a preconceived opinion, and you'll likely get an answer that supports it. The fact that usage runs above average among people on the political fringes makes this especially dangerous. Chatbots could deepen polarization rather than counter it. The second risk is further fragmentation of public discourse. The highly personalized way chatbots present news accelerates a trend social media already started. When every user gets a version of the news tailored to their interests, reading level, and preferences, the shared information base that public debate depends on erodes. Personalization could also broaden perspectives At the same time, chatbot personalization has an upside. It can make news more accessible to people who don't connect with traditional formats by simplifying complex topics, translating content into a user's preferred language 33 percent of users do this, according to the study , or adapting to individual information needs. Used well, AI chatbots could even expose people to a wider range of viewpoints. The study found that 35 percent of users say they use chatbots to pull together reports from multiple media sources. Those who actively seek out different perspectives can potentially get a broader picture than they would from any single news outlet. AI News Without the Hype – Curated by Humans Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section. Subscribe now Reuters Institute https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026/emerging-uses-ai-chatbots-news-and-what-it-means-journalism