One person in ten now uses an AI chatbot every week to get the news. That share jumped from 7 percent to 10 percent in a year, with young adults leading the move. Trust in what these chatbots say about the news stays low, even as more people rely on them. The shift is small in numbers but loud in direction.
The Gist
- Weekly AI chatbot use for news rose from 7 to 10 percent worldwide.
- 17 percent of people aged 18 to 24 already get news from chatbots; only 5 percent of older adults do.
- Only 20 percent of the general public trusts chatbot news, against 37 percent for traditional news.
Have ChatGPT Recap This Article
ChatGPTWhat the numbers actually say
The big number comes out of the Reuters Institute Digital News Report 2026. It is one of the most cited yearly studies on how people consume news. The headline this year: weekly use of AI chatbots for news climbed from 7 percent to 10 percent globally, in just twelve months.
A chatbot here means a tool like ChatGPT, Gemini or Claude. These are systems that talk to you in plain language and can summarize, explain, or pull in fresh information when connected to the web. They were not built to be news apps. People are using them as such anyway.
Only 1 percent of respondents call an AI chatbot their main source of news. So the 10 percent figure is not “people who replaced their news app with ChatGPT.” It is people who, in a given week, ask a chatbot a question that touches on current events. The two behaviors are very different, and the gap matters.
Where the growth happens is striking. The biggest gains landed in Asia, Africa, Latin America, and Southern and Eastern Europe. The United States and Northern Europe moved less. The pattern echoes other recent surveys showing falling AI optimism in the US while emerging markets keep pushing adoption forward.
Who actually does it, and why
Age is the strongest split. People aged 18 to 24 use chatbots for news at 17 percent. The oldest age group sits at 5 percent. The 25 to 34 bracket is moving fastest, with the strongest year-over-year jump. That is the same pattern we are seeing in how parents adopt AI to help raise their kids: younger adults move first, the rest follow with a delay.
Political views matter too. People with far-left views report 16 percent weekly use. Far-right views report 15 percent. The middle of the spectrum sits lower. The pattern hints that what brings people to AI for news is engagement with politics, not the side of politics they sit on.
“News lovers”, as the report calls them, are also heavier users at 18 percent, against 7 percent for casual news consumers. So the chatbot is not yet a tool for people who barely follow the news. It is, for now, an extra layer for those who already check the news in other places and want more depth, more speed, or a quick summary.
What do people actually ask? The top use case is the follow-up question, at 42 percent. People read or watch a piece somewhere, then ask the chatbot to explain a term, a context, or a related event. Getting current news comes second at 35 percent. Summaries land at 34 percent. Checking whether a source is reliable, 33 percent. Simplifying complex stories, 30 percent.
Notice what is not on the top list: replacing a news subscription. The chatbot is acting like an assistant next to the news, not as a news source on its own. That matters for media companies trying to decide whether AI is friend, foe, or both.
Keep learning on AI Noobies:
- China just cut 12,000 college majors because of AI
- 79% of parents now use AI to raise their kids
- Americans AI optimism collapses: only 16% left
Why trust stays so low
The trust numbers are the most awkward part of the report. Only 20 percent of the general public says they trust the news they get from AI chatbots. Traditional news still pulls 37 percent. So the gap is wide. People use the tool more, and still mistrust what it tells them.
Among active chatbot users, trust climbs to 44 percent. Non-users sit at 17 percent. The more you use these tools, the more you trust them. That is reassuring and worrying at the same time. Reassuring because the experience seems good enough to build confidence. Worrying because chatbots still make confident mistakes and people may not learn to spot them.
Source checking is the soft underbelly. Only 4 percent of respondents say they always or often click through to original sources when a chatbot answers them. Search engines do better at 19 percent. Social media at 17 percent. So the chatbot wins on convenience, but loses on the basic habit of reading the source.
Short term, the impact for regular users is simple. If you start asking a chatbot about the news, get into the habit of asking for sources explicitly, and click at least one of them. A chatbot that knows you check will tend to ground its answers more often. A chatbot that thinks you never check has no reason to be careful with you.
Medium term, the gap between use and trust is going to push news organizations and AI labs to talk to each other a lot more. Licensing deals, citation standards, and clear source labels are no longer optional. The 10 percent of today will be 20 percent tomorrow, and the way the chatbot picks its sources will shape how a generation forms its view of the world.
Follow the story on AI Noobies.
About the Author
