NYT, Economist, AFP join SPUR coalition to curb unpaid AI use of publisher content

The New York Times, The Economist and Agence France-Presse have joined a growing coalition of publishers seeking greater control over how artificial intelligence companies use journalistic content for training, retrieval and AI-generated responses.

More than 40 publishers joined the Standards for Publisher Usage Rights (SPUR) coalition in its latest expansion, taking membership to over 80 organisations across more than 20 countries. New members also include The Atlantic and Brazil’s National Association of Newspapers, while founding participants include the Financial Times, BBC and Associated Press.

Read On: OpenAI, Microsoft ‘scraping’ content without permission: Publishers

SPUR is seeking to create common technical and commercial standards that allow publishers to monitor how their content is being used by generative AI systems and, where they choose to license it, enable compensation and transparent usage reporting. 

Earlier this month, the coalition released the first version of its Content Telemetry Standard, an open-source framework designed to record when publisher content is retrieved, used to ground an AI answer, cited, displayed or engaged with through an AI platform.

The initiative comes as tensions between publishers and AI companies escalate over copyright and compensation. Several media companies have pursued legal action alleging that copyrighted journalism was used without permission to train large language models, while publishers have also raised concerns about AI-generated answers reducing referral traffic to news websites.

Read On: Publishers sue Google, allege Gemini was trained on copyrighted books without permission

SPUR, however, is not a collective licensing body. Publishers retain control over their individual commercial agreements, with the coalition focused instead on building common infrastructure and standards around permissions, measurement and licensing.

In India, the debate is playing out through the ANI-OpenAI copyright case. The Delhi High Court ruled in July that OpenAI’s use of ANI content for model training qualified as fair dealing, although ANI has appealed the decision and the court sought OpenAI’s response in September. The case is being closely watched for how Indian copyright law will ultimately treat AI training on publisher content.