One in ten people now get their news from an AI chatbot

Ten percent of people worldwide now use an AI chatbot for news at least once a week. A year ago it was seven percent. That is the headline number from the Reuters Institute's Digital News Report 2026, the biggest annual survey of how people actually consume news, and it is the fastest-moving line in the whole report.

Three points in a year sounds small. It isn't. It is the difference between a habit a few early adopters have and one a tenth of the planet has, and the curve is still pointing up.

Who is actually doing this

Age splits it hard. Among under-35s, 17% use chatbots for news each week. Among the oldest group surveyed, it is 5%. So the youngest readers are roughly three times more likely to ask ChatGPT or Gemini "what happened today" than their parents are.

Geography splits it just as hard. Weekly use hits 14% in South Korea and Turkey and 13% in Brazil. It drops to 6% in the United States and just 4% in the United Kingdom. The growth is loudest across Asia, Latin America, Africa and parts of Southern and Eastern Europe, where news already lives inside big platforms rather than on newspaper front pages.

Why are people doing it? The report's answers are boringly practical. Chatbots are fast. They summarise a tangled story in a paragraph. You can ask a follow-up, or get it translated, without opening five tabs.

The part publishers are sweating over

Here is the number that matters if you make a living writing the news: only 42% of chatbot users say they always or often click through to the original source. The other 58% read the chatbot's summary and move on.

That breaks the deal the open web ran on for twenty years. Search sent you a link, you clicked, the publisher got a visit and maybe an ad view. A chatbot answers the question on its own page, often using reporting it scraped from sites that now get nothing back. One industry analysis of the report put it bluntly: the audience is being eaten by social feeds and chatbots at the same time.

Trust is the other catch. Globally, only 20% of people say they trust news that comes from AI chatbots. The problem is that a chatbot answer reads clean and confident even when it is wrong, which is exactly the tone you do not want wrapped around a made-up fact.

What it means

Don't read this as "chatbots are replacing the news." Just 1% of people say a chatbot is their main source. For almost everyone it is a top-up, a quick "catch me up" they use alongside everything else.

But 1 in 10 weekly, doubling-ish in the youngest group, with most people never clicking through, is a real shift in who controls the front door to the news. The story to watch over the next year isn't the percentage going up. It is whether anyone is left to write the news the chatbots summarise.