AI is becoming an information gateway. Audiences still don’t want it making their news.

In my Nieman Lab prediction this time last year, I wrote that “in 2026, the use of AI as an information gateway will only rise further.”

As it turns out, I was right. In new research that I published with my colleagues Rasmus Kleis Nielsen and Richard Fletcher, one of our many findings is that generative AI systems are becoming ever more significant information intermediaries, both when intentionally used and when incidentally encountered.

For our new report, “Generative AI and News Report 2026: From Standalone Tools to Information Infrastructure,”  we surveyed audiences across nine countries, including the United States, on their use of AI in their everyday lives, what they think its impact will be on different sectors of society, and what they think about its use in news and journalism specifically. I obviously think you should read the whole report, but here are some of the major takeaways.

The use of standalone conversational AI systems such as ChatGPT, Claude, and Gemini continues to grow rapidly, with 51% of respondents across the six countries where we have past data saying they now use AI weekly (49% in the U.S.). That’s a faster spread, at least in the U.S., than smartphones or the internet saw.

Many people also encounter AI through “incidental” AI exposure in the AI-infused services and platforms they’re already using. Search engines such as Google and Bing are key drivers: 57% of respondents across countries said they saw or engaged with AI-generated answers or summaries in the last week. News media, too, drives AI exposure. An average of 24% of respondents across nine countries said they’d seen a label noting that content was made with AI in the last week, and 20% had seen a bullet-point AI news summary.

Despite all the talk about AI agents, we find that, so far, they haven’t broken into the mainstream. The vast majority of respondents across all nine countries (67%) said they had not used an AI agent in the last week. Fourteen percent used one for a real task and 13% tried one just to experiment.

Conversational AI as a friction-reducing information intermediary

Our data provides solid evidence that a broader transformation of the information environment is afoot. Conversational AI systems have arrived as information intermediaries. Trust is not the main reason people use them. Instead, people see conversational AI systems as friction-reducing machines that get them information more quickly.

“Getting information” continues to be the most common use case. Thirty-eight percent of respondents across countries (35% in the U.S.) said they had used generative AI for getting information in the last week, up from 24% in 2025 and 11% in 2024.

We looked more closely at what people do when they say they use AI for getting information. Across the six countries we’ve surveyed since 2024, we primarily see a mixture of search-equivalent behaviors like “researching a topic” (24%) or “answering factual questions” (18%). People are also increasingly turning to AI for advice (20%).

Using AI directly for news, meanwhile, remains a niche activity, as we demonstrated in the Digital News Report earlier this year. But that use is increasing: 9% of respondents across our six core countries say they use AI for news, up from 6% in 2025 and 3% in 2024.

Many people don’t hold clear or strong views about whether AI output is trustworthy. But people who do rely on it for information trust it. Across the nine countries, 75% of those who used AI to answer factual questions trusted the answers, as did 76% of those who used it for advice and 72% of those who used it for news. Only small minorities expressed distrust.

At the same time, trust and trustworthiness do not, at this stage, seem to be central to why people rely on AI systems. Instead, people tend to see these systems as friction-reducing machines. The convenience of getting a faster and easier answer was chosen by 64% as the main reason they use AI, followed by the interactivity afforded by natural language interfaces (56%). Only a minority (21%) said they use AI because they get more reliable information from it than from other sources.

As AI systems become widely used, their growing popularity is likely to affect the use of legacy content offerings and digital platforms and services. As my coauthor Rasmus Kleis Nielsen put it on LinkedIn, rapid growth in AI use isn’t just eating into what news sites have to offer. It’s eating into search, too — with people who use AI for information saying they use not just news, but search engines less, too, and by a wider margin. Challenging times may be ahead not just for news publishers, but also for the incumbent digital giants previously seen as unassailable.

Persistent attitudes: People’s views on generative AI in journalism and news

As in previous years, we wanted to know what people think of AI’s use in journalism and news. We find that people’s attitudes here are remarkably entrenched and negative — though they also haven’t worsened significantly.

People across countries remain uncomfortable with the use of AI in the newsroom, especially with the idea of news being made entirely or primarily with AI. On average across nine countries, just 12% of respondents say they are very or somewhat comfortable with news made entirely by AI. That figure rises to 19% if there is some human oversight. This has hardly changed since 2024, and the gap between human- and AI-made news appears in every country we surveyed.

This also remains true regardless of whether it is breaking news or reporting. People are even more uncomfortable with AI being used for commentary or opinion — formats where the expectation that a human should share their view is strongest.

Zooming out

Our findings raise several important questions for journalists and publishers navigating the rise of generative AI.

First, generative AI is becoming an increasingly significant information intermediaries, with a number of possible knock-on effects on news. People are using AI to reduce friction by getting “good enough” outputs for some information needs — and skipping legacy platforms, established publishers, and the time and effort it takes to find the best available information. Discovery, brand attribution, data collection, and monetization all grow more challenging for publishers in this context.

Second, even though use continues to grow, the public remains skeptical of AI in news. This isn’t just a challenge for newsrooms’ internal adoption. It also shows that publishers using AI need to communicate to audiences what’s in it for them.

Third, beyond news, our findings speak to wider debates about AI’s societal impact. As in other studies on this topic, we find that uptake of AI is uneven. Weekly use is more common among people in professional occupations than those in entry-level or service occupations. Most people rely on free — usually less advanced — AI models.

The differences in how — not just whether — people use AI may become a new dividing line. As AI researcher and OpenAI cofounder Andrej Karpathy recently argued, there is a widening “gap in shared understanding of LLM capability,” with most people not seeing what has left those inside AI firms both excited and worried.

Expectations of what AI will do differ far more between daily users and people who have never used it than between any other groups in our survey. When two groups form very different views of the same technology because they have such different starting points, there will likely be consequences.

This doesn’t matter just for people’s individual choices around whether and how to make use of AI in their own lives. It also matters where AI is adopted and governed through collective decisions — in public services, for example — that apply to those who do not use it.