The Scarce Resource

The Neal Conan Prize
Jul 25
3 min read

Updated: Aug 1

A month later, one number is still with us.

Last month we wrote about the 2026 Digital News Report and the growing distance between reporting and the people who depend on it. Platforms and chatbots increasingly stand in the middle. That shift is easy to describe as another loss for journalism. It is more useful to ask what becomes newly valuable because of it.

On July 7, the Reuters Institute returned to the report’s findings in a conversation about how people use AI chatbots for news. Ten percent of respondents worldwide said they had used one for news in the previous week, up from 7% a year earlier. Only 1% called a chatbot their main source. The adoption is still uneven and concentrated among people already interested in news, but the path from question to answer is plainly changing.

Across the 27 markets where the Institute asked about click-throughs, just 4% of respondents said they often or always followed a chatbot answer to an underlying news source. That compares with 19% for search and 17% for social media. The figures are self-reported, and the lower rate partly reflects the fact that chatbot use remains far smaller than search or social use. Still, the audience experience matters: an answer can feel complete even when the reporting that made it possible has disappeared from view.

Answers without the reporting

People are not only asking chatbots for headlines. They use them to pose follow-up questions, simplify complicated news, summarize it, and even evaluate whether a source is reliable. Those are acts of interpretation—the work between a fact and an audience’s understanding of it.

Journalists have always done that work in public. A byline, a newsroom, a correction, a document, and a visible chain of evidence give readers something to test. A fluent answer with no obvious provenance offers a different bargain. It may be useful. It may also make authorship, accountability, and the cost of the reporting difficult to see.

When the answer is cheap and instant, the scarce resource is not more language. It is judgment: about the source, the reporter, the evidence, and the work worth believing.

The terms side of the question

That is why the developing work of the SPUR coalition belongs in the same conversation. SPUR is proposing systems for publishers to express permission, payment, and provenance as journalism moves through generative-AI systems. Its focus is infrastructure: opt-in licensing, transparent data flows, and a fair exchange of value.

It is a proposal, not a proven solution. Standards matter only if powerful companies honor them, and a framework designed around large publishers could leave smaller newsrooms with little leverage. The test is not whether the language sounds fair. It is whether reporters and publishers of different sizes can control the terms on which their work enters the machines—and whether audiences can still find the reporting underneath the answer.

Distribution can be automated. Trust cannot.

The choice before us

The Neal Conan Prize cannot redesign the information economy. It can make one narrower act of judgment visible. Each year, it identifies a mid-career journalist whose rigor, curiosity, humanity, courage, nuance, and respect for the audience deserve material support: $50,000, unrestricted, for the work still to come.

That choice matters more in a system that can reproduce language at enormous scale. Distribution can be automated. Trust cannot. It is earned by people whose names are attached to the work and whose methods can withstand a question.

Nominations for the 2026 Neal Conan Prize close July 31. If one journalist comes to mind, email info@nealconanprize.org with their name. One name from you could shape the next chapter.