Unsealed court documents in the New York Times lawsuit against OpenAI and Microsoft show the companies had internal warnings before they acted. Their own documentation flagged a doom loop that would damage the open web, described their training data scraping as the largest theft of labor in human history, and said it made a complete mockery of the idea of fair use. They built anyway.

The Verge reports that the most pointed quotes come from Microsoft's Director of Applied Science, Brent Hecht. Microsoft has since tried to distance itself, with spokesperson Alex Haurek stating the comments reflect one employee's views. That distancing move is notable: internal documentation and a named director-level source are harder to dismiss than a leaked memo.

The operating question now is liability. If internal documentation shows companies knew the harm and proceeded, fair use defenses weaken. Publishers and regulators should watch how courts treat the doom loop framing, because that language came from inside the house. This is not a case of an external critic raising alarms. It is the defendants' own record.