Reuters Institute: 1 in 10 Adults Now Use AI Chatbots for News Weekly—But Trust Sits at Just 20%

The Reuters Institute for the Study of Journalism released its fourteenth annual Digital News Report last week, and for the first time in the survey’s history, it describes the global news environment as entering what the researchers call “unsettled time.” The headline numbers are striking: 10% of global adults now use AI chatbots for news at least once a week, social media has officially overtaken all other sources as the most widely used news channel, and trust in AI-delivered news sits at just 20%—barely half the 37% trust in journalism overall.

The report surveyed more than 90,000 people across 48 markets between January and February 2026. Its findings arrive at a moment when AI labs are investing heavily in news distribution partnerships, chatbot-native news formats are proliferating, and traditional newsrooms face accelerating pressure to define their relationship with AI-generated content.

The Growth of AI as a News Source

Weekly AI chatbot use for news rose three percentage points over the year—from 7% to 10% globally—driven largely by growth in parts of Asia, Africa, and Latin America, as well as Southern and Eastern Europe. Only 1% of respondents say AI chatbots are their main source of news, suggesting AI currently plays a supplementary rather than primary role in most people’s news diets.

The growth, however, is not uniform. South Korea, Greece, and Spain each approximately doubled their year-on-year adoption rates. Major English-speaking markets told a different story: the USA, UK, France, and Germany reported no meaningful increase in AI chatbot use for news over the year. The UK records the lowest adoption rate globally at just 4%.

Demographics remain the dominant predictor of adoption. Among respondents under 35, AI chatbot use for news stands at 16%—60% above the global average. “Uptake continues to be driven by younger people,” the report states, with the youngest cohort’s usage rate approximately three times that of the oldest. Those who use AI for news also skew toward “news lovers”: 38% of AI chatbot news users fall into the high-engagement segment, compared to 22% of the overall sample.

What People Actually Do with AI News

The most-valued feature of AI chatbots for news is straightforward: 42% of AI news users cite “the ability to ask follow-up questions” as their primary use case. Roughly a third use AI for summarization (34%) or to make difficult news easier to understand (30%). About a third (33%) use chatbots to assess the reliability of a news source or story—an ironic use case given how poorly AI rates on the trust metrics.

The click-through question has been a persistent anxiety for publishers: if AI summarizes the news, do readers ever visit the original source? The 2026 data provides modest reassurance. Approximately 42% of AI chatbot users say they “always or often” click through from a chatbot answer to the original news source. That proportion varies significantly by market—from 26% at the low end to 56% in the most engaged markets—and skews toward younger, higher-education respondents.

Whether 42% click-through is adequate from a journalism business-model perspective is a different question. Traditional search delivered far higher referral rates, and the economics of AI-mediated news distribution remain unresolved at every major publisher.

Social Media Crosses a Landmark

The bigger structural shift in the 2026 report may not be AI at all, but social media’s formal crossing of a significant threshold. For the first time, the share of adults using social media and video networks as their primary news source globally (54%) exceeds all other channels combined.

Among 18-to-24-year-olds, 52% now say social media, video networks, and AI chatbots collectively represent their main way of getting news. That cohort is the audience that journalism schools, newsrooms, and media companies have spent a decade trying to reach, and they have largely redirected their attention to platforms that were not primarily designed to inform them.

The dynamics are self-reinforcing. As social media becomes the dominant news environment, creators who understand social platforms—rather than journalists who understand editorial standards—become the most influential voices. The report documents a significant increase in news creators (what the Reuters Institute calls “the creator economy”) displacing traditional news brands, particularly on YouTube, TikTok, and Instagram.

The Trust Deficit

The trust numbers are a recurring theme across the entire report, and AI chatbots fare the worst. Trust in news from AI chatbots globally: 20%. Trust in news from social media: 22%. Trust in news overall: 37%. Trust in news “I use myself”: approximately 44%.

The pattern is consistent: the newer and more algorithmically mediated the source, the lower the trust. The UK presents the extreme case—6% trust in AI chatbot news, the lowest across all surveyed markets, and an AI adoption rate of just 4%.

Researchers caution against reading the low trust numbers as evidence that audiences are rejecting AI news. The data shows low trust alongside consistent or growing usage—a combination the report describes as reflecting “pragmatic adaptation rather than endorsement.” People use tools that provide utility even when they don’t fully trust them; the same dynamic has characterized social media use for years.

Implications for Newsrooms

The report stops short of prescriptive recommendations, but its findings have clear implications for news organizations navigating the AI transition.

The click-through data suggests that AI-mediated news does not automatically eliminate the referral relationship between chatbots and publishers—but it attenuates it, concentrates it among the most engaged users, and makes it contingent on the chatbot choosing to surface source links prominently. That last point is an editorial and commercial decision made by AI labs, not publishers.

The use case data—follow-up questions, summarization, comprehension assistance—points toward a hybrid model where AI handles initial orientation and traditional reporting handles depth. The journalists who will matter most in an AI-mediated news environment, this reading suggests, are those who produce original reporting and analysis that AI cannot replicate: ground-level observation, source cultivation, investigative work, narrative journalism. Commodity news—who won the game, what the stock did, what the official said—is already being commoditized.

What “Unsettled Time” Means

The Reuters Institute’s framing of this moment as “unsettled” is deliberate. The report tracks neither a clear trajectory toward AI as the dominant news mechanism nor a rejection of it. It documents an industry and an audience in transition, trying to establish new norms for how information flows, who is trusted to provide it, and what obligation exists to original reporting.

What is clear from the data is that the transition is happening faster among younger, more urban, more globally connected audiences—and slower in the traditional strongholds of quality journalism: Germany, the Nordic countries, Japan. Whether the gap between those cohorts closes toward the chatbot end of the spectrum or stabilizes at a new equilibrium is the defining uncertainty for the next five years of global journalism.