AI in news: "Need to Ask” vs “No Need to Ask

This September, I attended the European Business Media conference in Portugal. The network brings together editors, publishers, and media executives from across Europe to share insights and navigate changing tech and business models.

Almost every media house took the stage to share their latest work, including their successes, hard-learned failures, and new revenue models. Unsurprisingly, this year’s presentations were dominated by AI in news.

What made the sessions so interesting was the variety of approaches. Some publishers focused purely on internal tools to help their newsrooms work faster, while others built features directly for subscribers to boost engagement. Many did both. Seeing these different experiments side by side highlighted just how much potential is still untapped, which is exactly why getting together at events like this is so valuable.

AI in news: chat tools vs personalized news feeds

What stood out to me most was a clear divide in how publishers approach reader-facing AI features: “Need to Ask” versus “No Need to Ask”.

The “Need to Ask” approach relies on interactive chat widgets or search boxes. As a reader, you type in a question, and the tool builds an answer drawn directly from that publisher’s archives.
On the flip side, the “No Need to Ask” approach relies on background personalization. By looking at what you have read in the past, the site automatically surfaces relevant articles, tailored homepages, and suggestions without you ever having to search for a thing.
Personalization itself is, of course, not new for publishers – as the Reuters Institute explains, AI is mainly opening up new possibilities for tailoring both the selection and presentation of news. 

That contrast got me thinking. When we open a news site, do we actually want to have a conversation with the content, or do we just want to read it?

The reality is that typing prompts takes mental effort, and most of us visit news sites during quick breaks or on our morning commute when we are running low on energy. We often do not even know the exact questions we ought to ask. Publishers already have decades of quality reporting and plenty of data on what their audience cares about, so in many cases, they already know what a reader needs before anyone touches a keyboard.

Different reader needs call for different approaches

Reflecting on the general approach of AI in news on the plane ride home, my takeaway is that there is room for both approaches.

How we want our news depends heavily on our mindset at that specific moment. A retail investor who just wants a quick morning market update does not want to sit there typing out queries. They want a clean, personalized summary waiting for them the second they open the app. But someone doing deep research on a specific company needs to dig into the details, and for them, an interactive Q&A tool tied to verified archives is a huge time-saver.

Trust matters in both approaches

At the end of the day, whichever method publishers choose, the success of AI in news  comes down to trust. Readers have to know that the information in front of them is accurate and reliable, whether an algorithm selected it or a chat tool summarized it.

I will dive deeper into why trust is the real make-or-break factor for publisher AI in my next post.

When you open your favorite news app, which do you prefer: a feed tailored to your interests, or a tool you can actually talk to?