What the 2025 Digital News Report Tells Us About AI in News

Based on responses from over 100,000 people across 48 markets, the 2025 edition of the Reuters Institute Digital News Report offers one of the first large-scale glimpses into how global audiences perceive and engage with AI in news.

This article explores three key takeaways from the report and what they mean for publishers navigating AI integration in the newsroom.

1. Audience interest in AI is emerging. Now is the time to experiment broadly.

The chart below illustrates the percentage of respondents who expressed interest in various AI-powered news personalization features.

At first glance, the picture looks lukewarm. No single feature scored above 30% interest. It would be easy to interpret this as a lack of appetite for AI in news, but that would miss a critical nuance: The report also finds that 66% of respondents expressed interest in at least one AI application. This shows that there is a widespread interest in incorporating AI into news experiences, but that the jury is still out on exactly which one(s) would be of most value.

What should publishers take away?

Now is the time for publishers to explore a wide range of AI-driven features to see what truly resonates with audiences. But experimentation must be thoughtful. Understanding and mitigating why audiences may hesitate is essential during this stage (e.g., due to trust, privacy, or relevance concerns).

2. Chatbots: A new channel, still far from mainstream.

AI chatbots (e.g., ChatGPT, Gemini, custom news ones, etc.) are beginning to enter the news ecosystem, but for now, they remain a relatively niche source of news. Only 7% of respondents report using a chatbot for news on a weekly basis. Among younger audiences (under 25), however, this figure rises to 15%, suggesting a generational gap in adoption.

There is an ongoing debate about why adoption isn’t higher yet. A few discussed hypotheses include:

• Active engagement reduces interest in use: Chatbots demand user input, unlike the passive feed-based experiences most users are accustomed to
• Lack of brand presence: Users increasingly value trusted journalistic voices or personalities, which chatbots lack
• Trust concerns: Many users remain sceptical of the reliability and accuracy of chatbot-generated content
• Unpolished UX: Current interfaces are not yet tailored for intuitive news discovery and consumption

While chatbot use remains relatively low today, the upcoming rollout of AI Mode from Google could rapidly increase exposure to AI-generated news, whether audiences actively seek it or not.

What should publishers take away?

At this stage, chatbots represent more potential than proven demand. They may evolve into a larger distribution channel, but for now, they remain one part of a broader AI experimentation landscape. While younger users are beginning to engage, mainstream audiences still require clearer value, stronger trust, and better UX.

3. Format adaptation is where audience interest and publisher action align.

The Digital News Report showed that audiences are more open to AI features that help them consume news more efficiently, rather than those that choose news for them.

The graph below shows that the most popular AI use cases were article summaries (27%) and translations into other languages (24%). These are practical, low-friction features that enhance access and convenience, particularly for time-constrained readers or those in small linguistic markets looking to read news from abroad.

The graph also shows how publishers are responding, mapping audience interest against where media leaders say they plan to invest in 2025. We clearly see that media leaders are mainly prioritizing low-risk, high-utility applications like summarization, audio conversion, and translation. These are relatively easier to implement and present fewer editorial concerns than more complex or controversial features, like chatbots (although 56% of publishers are still planning to invest in this feature).

What should publishers take away?

Audiences are asking for support in how they read, listen, or watch. This is an encouraging signal. It means there’s space to experiment with AI in a way that adds value without threatening trust. We’re still in the exploratory phase. Starting with high-demand, lower-risk features makes sense. But as these tools mature, publishers should stay close to user needs and push beyond the basics.

Conclusion

This year’s Digital News Report offers a valuable early snapshot of how audiences are beginning to engage with AI in the news. While adoption remains limited and trust is still developing, there are clear signs of curiosity.

But AI is only one piece of a much larger puzzle. The report also sheds light on the growing fragmentation of platforms, deepening generational divides in consumption habits, the rise of influencers as news vectors, and ongoing concerns around trust and credibility in journalism.

For anyone working in product, audience, or editorial strategy, this report is essential reading. AI may be one of the more eye-catching trends, but it’s the interplay of all these forces that will shape the future of news.