NYT, AFP and The Economist Join the SPUR Coalition, Now 80+ Publishers Strong

The New York Times, Agence France-Presse and The Economist have joined SPUR (Standards for Publisher Usage Rights), a coalition that wants AI companies to stop using news without permission. The group announced the move on Monday. The coalition was founded in February and now has more than 80 members from more than 20 countries. For an organization that is not yet a year old, that is fast growth. What matters most is that the newcomers have very different track records with the AI industry.

Three Different Publishers, One Position

NYT has been suing OpenAI and Microsoft since December 2023 over the training of models on its articles. At the same time, the newspaper has signed a licensing agreement with Amazon, so it is not opposed to cooperation on principle. AFP has also gone the deal route: in early 2025, the agency reached an agreement with Mistral to use its content in the Le Chat chatbot. The Economist represents a third type of publisher: a global subscription title for which analysis is the core product.

The underlying logic is simple. Publishers are not against AI as such. They want access to their content to be governed by their permission, with terms set by contract. That word, "permission," is at the center of SPUR's statement.

What We Know About the Coalition

The available part of the announcement gives no details about the coalition's legal form, what the standards will contain, or how it plans to get them enforced. So it is best to stick to the confirmed facts:

• Name: SPUR, Standards for Publisher Usage Rights.
• Founded: February of this year.
• Membership: more than 80 members.
• Geography: more than 20 countries.
• New members: The New York Times, Agence France-Presse, The Economist.
• Main demand: AI companies do not use news without publishers' permission.

The word "standards" in the name points to an attempt to agree on common rules instead of conducting dozens of separate negotiations. For AI developers, that may be more convenient than lawsuits. But only if the rules turn out to be technically clear.

Where the Stress Test Comes

The industry already has tools that try to solve the same problem: robots.txt, paid crawler-access schemes such as Cloudflare's pay-per-crawl, and licensing standards such as RSL. None of them has become universally accepted yet. A new standard will work only if the model developers themselves support it. Otherwise it will remain a declaration, even if it is signed by big names.

The second test concerns the difference between training and search. Using news to train a model and showing it in an agent's answer in real time are legally and economically different situations. The short announcement does not make clear which scenario SPUR has in mind. The answer determines how the coalition will affect products built on search and RAG.

What Teams Should Do

Teams building products on news data, whether search assistants, research agents or market monitoring, should check the provenance of their content in advance. They need to know which sources end up in the index, on what terms, and whether they have written permission. It helps to keep a source registry that records usage rights: when standards do arrive, the system will need far less rebuilding.

For publishers that have not yet made up their minds, the coalition offers a ready-made platform for negotiations. Watch whether SPUR publishes technical specifications, and treat that as the main measure of how serious the project is.