Reuters Institute: AI use hits 51% weekly across nine countries

Weekly generative AI use across six tracked countries rose from 34% in 2025 to 51% in 2026, and daily use nearly doubled year-on-year from 16% to 28%, according to the Reuters Institute's Generative AI and News Report 2026, published on 29 October 2026.

The study, led by Dr Felix Simon, Prof. Rasmus Kleis Nielsen and Dr Richard Fletcher, surveyed around 2,000 people each in Argentina, Australia, Denmark, Finland, France, Japan, Norway, the UK and the US via YouGov between 14 August and 28 September 2026.

Information-seeking has overtaken every other use case. Getting information rose from 11% in 2024 to 38% in 2026. 24% use AI for researching topics, 20% for advice, 18% for factual questions, and 9% for news, up from 3% two years ago. 57% say they come across AI-generated summaries in search engines weekly.

Public comfort with AI-made journalism has barely moved. Only 12% are comfortable with entirely AI-generated news, 19% if a human oversees it, and 67% with entirely human-written news. The authors describe these attitudes as "remarkably sticky" and largely unchanged since 2024.

Trust in the assistants themselves runs higher than trust in the organisations using them. ChatGPT is trusted by 37% and leads weekly use at 34%; Google Gemini is trusted by 27% and has gained 13 points of weekly use to reach 24%; Microsoft's Copilot is trusted by 18% and Anthropic's Claude by 13%. Of users who ask AI factual questions, 75% trust the answers; 72% trust AI news summaries.

"Across all sectors, there is an honesty problem when it comes to AI," the report concludes, with public optimism skewed toward scientists and healthcare workers and pessimism toward politicians, governments and news media.