When you ask an AI assistant what happened in the world yesterday, where does its answer come from? Very often, from articles written by journalists.
Those journalists, and the newspapers that employ them, are mostly not paid for it. More than 80 news publishers have just decided to join forces to change that.
What happened
On Monday, Agence France-Presse, the New York Times and The Economist joined a coalition called SPUR, for "Standards for Publisher Usage Rights". They are among more than 40 new members, including The Atlantic, the British broadcaster ITV, and two French outlets: Radio France and L'Équipe.
The coalition now has more than 80 publishers in over 20 countries. It was launched in February in London by the BBC, the Financial Times, the Guardian, the Telegraph and Sky News, joined in July by the Associated Press. It is entirely funded by its members.
A different strategy from lawsuits
Until now, each newspaper fought alone, often in court. The New York Times is suing OpenAI and Microsoft for copyright infringement, and the US group USA Today has also just taken action against OpenAI.
SPUR is betting on something else: a shared tool. In early October, the coalition published the first version of a technical standard designed to track how AIs retrieve, use, cite and display news articles. In other words, a meter. Because to be paid, you first have to be able to measure what was used.
According to the specialist site Digiday, the coalition has invited OpenAI, Anthropic, Google, Meta and Microsoft to take part in an advisory committee on licensing. So the door remains open to negotiation. A New York Times executive summed it up this way: protecting intellectual property and advancing technology can and must go together.
In 2019, France was the first country to apply the European "neighbouring rights", which allow news publishers to be paid for the reuse of their content online. And in March 2024, the Autorité de la concurrence fined Google 250 million euros, notably accusing it of using press content to train its AI assistant of the time, Bard, without informing publishers. A precedent that European media have in mind.
Why it matters
The stakes go beyond money. If you get most of your information from an AI answer, you no longer click on the original article. Newspapers then lose their readers, and the revenue that goes with them.
Yet without journalists to investigate, verify and report, AIs would have nothing reliable left to summarise. It is a circle that can turn vicious: AI lives off the press, but can weaken it. Hence the importance of always tracing information back to its source when an AI gives it to you.
It is the same question raised by British universities about student essays used by AI: who has the right to use other people's work, and at what price?
What to take away
Publishers have understood that they would carry more weight together than apart. With AFP and the New York Times, the coalition now has names that the AI giants can hardly ignore.
In the age of AI, information is not free. It has only been free for those who reuse it. The real question now is who will pay, and how much.