Sources: OpenAI, Google, Microsoft, and Adobe have met News Corp, the NYT, and others to discuss copyright issues over LLMs and a possible subscription fee
Context & Ripple Effects
This meeting is the opening move in what became the defining fight over training data: weeks later News Corp, the NYT and IAC would go further, forming a coalition to push for AI laws and weigh a lawsuit, while by year-end OpenAI was already deep in pricing-and-terms negotiations with Gannett, News Corp and IAC.
The FT's detail that a subscription fee was on the table matters because it shows the labs treating news content as a purchasable input from the start — even as Microsoft and OpenAI would later defend the same scraping in federal court against dismissal demands from the outlets.
First-order effects
- News Corp and the NYT move from passive scraped suppliers to negotiating counterparties, with Adobe's presence signaling that the buyer side extends beyond chatbot makers into creative software.
- OpenAI, Google and Microsoft get a path to licensed training corpora that could de-risk the copyright exposure they were simultaneously fighting over in court.
Second-order effects
- If a fee structure emerges, rival buyers follow suit rather than risk being the last lab without licensed text — Meta's later approach to Axel Springer, Fox Corp and News Corp for its own AI tools shows exactly that copycat dynamic.
- Publishers split into two camps: those monetizing through negotiation and those litigating, as when Alden's eight dailies sued OpenAI and Microsoft outright — pressuring holdout labs on both fronts.
Third-order effects
- Licensed news content hardens into a recurring line item in model-development budgets, shifting economics so that whoever controls distribution and rights aggregation captures margin across the supply chain.
- The pattern points toward formalized content-as-inference-input markets, where collective publisher bargaining power — the coalition model — sets floor prices for training and retrieval access industry-wide.
The trend: News publishers are converting copyright leverage over LLM training data into paid licensing arrangements, turning editorial archives from free inputs into negotiated revenue streams.