Internal emails from OpenAI and Microsoft show executives acknowledging that AI products pose an 'existential threat' to publishers, with potential market impacts exceeding $10 billion. The revelations significantly weaken the companies' fair-use defense in The New York Times' lawsuit.
TL;DR
- OpenAI and Microsoft internal emails reveal executives acknowledging AI products as an 'existential threat' to publishers.
- The emails weaken the companies' fair-use defense in the NYT lawsuit by showing they knew AI could substitute publishers' content.
- The case could set a precedent for AI companies to pay for news content, strengthening publishers' bargaining power.
What happened
A new tranche of internal emails and analysis in The New York Times' lawsuit against OpenAI and Microsoft were unsealed last Thursday. The filings show executives discussing the 'gazillions' of dollars at stake in commercializing AI models while privately acknowledging the 'existential threat' these products pose to publishers' underlying economics.
The emails reveal that OpenAI and Microsoft executives were aware that AI products could substitute publishers' content, strengthening the Times' argument that the companies built AI products that can replace journalism rather than use it for a different purpose. This 'substitutive' use is a critical factor in fair-use cases.
The filings also show evidence of paywall circumvention, which puts OpenAI in potential violation of the Digital Millennium Copyright Act. This move is seen as a 'really big deal' both legally and morally, as it almost always defeats a fair-use defense.
Why it matters
The revelations significantly weaken OpenAI and Microsoft's fair-use defense, as the emails show they knew AI products could substitute publishers' content. This could hand publishers their strongest leverage yet over training data and licensing.
The case could open the door for more publishers to file lawsuits, pushing AI companies into a paid licensing market for news content. If the Times wins, it could set a precedent for all cases about training, affecting companies like Google, Anthropic, and Meta.
For publishers, the unsealed material validates years of frustration about how their work has been treated. It could force AI firms toward a paid market for news content, with publishers collectively negotiating and licensing their content at scale.
Key facts
- OpenAI and Microsoft executives acknowledged AI products pose an 'existential threat' to publishers, according to internal emails.
- The emails reveal that AI products could substitute publishers' content, weakening the companies' fair-use defense.
- Evidence of paywall circumvention puts OpenAI in potential violation of the Digital Millennium Copyright Act.
- The case could set a precedent for AI companies to pay for news content, affecting companies like Google, Anthropic, and Meta.
- Publishers could gain real bargaining power and more may file lawsuits if the Times wins.
- The unsealed material validates publishers' frustrations and could force AI firms toward a paid market for news content.
- Legal teams for the NYT, OpenAI, and Microsoft will challenge each other on which documents should be unredacted and released to the public.
- Both sides are asking Judge Sidney Stein for summary judgment, arguing that the evidence is clear enough to resolve key issues without a trial.
Context
This case is part of a broader conversation about the relationship between AI companies and publishers. AI models are trained on vast amounts of data, much of which is sourced from publishers' content. The use of this data without proper licensing or compensation has been a contentious issue.
The outcome of this case could have significant implications for the AI industry. If the court rules in favor of the Times, it could force AI companies to pay for the data they use to train their models. This could lead to a more structured licensing market, with publishers collectively negotiating and licensing their content.
The case also highlights the importance of fair-use laws in the digital age. As AI models become more advanced, the line between fair use and infringement becomes increasingly blurred. This case could help clarify these boundaries and set a precedent for future disputes.
