Paywall violation ‘eviscerates’ fair use defense: What OpenAI-NYT twist means for publishers

The filings, unsealed last Thursday, show executives talking about the “gazillions” of dollars at stake in commercializing AI models while privately acknowledging the “existential threat” those products pose to publishers’ underlying economics. 

It’s the clearest public window yet into what tech executives were saying internally about the issues at the heart of the case. And if the court buys this picture, it doesn’t just ding OpenAI and Microsoft; it reshapes the terms of trade between AI and publishing, and could hand publishers their strongest leverage yet over training data and licensing. 

Digiday spoke to a range of lawyers and AI licensing experts about what the filings signal for publishers’ AI monetization strategies. 

‘Eviscerates’ the fair-use argument for training on news 

Legal and regulatory experts say the newly unsealed documents are devastating for OpenAI and Microsoft’s fair-use claim over training on news.

The fourth fair-use factor is the impact on the market for the work. The filings show a real licensing market already exists and was operating inside the defendant’s partnership. Content was explicitly priced and traded between Microsoft and OpenAI, and all the major AI players now sign publisher licensing deals for similar uses, stressed David Buttle, founder of publisher AI licensing coalition Spur.