WorkSeptember 18, 2026via The Decoder
AI training built on fair use looks shaky when the companies' own people call it "astonishing theft"
Why it matters
A growing legal and reputational threat to foundational AI training practices. When a company's own leaders admit substitutive use and massive labor extraction in sworn testimony, fair-use claims collapse — and so does the cost model that made current AI economics viable.
Key signals
- Microsoft director called the practice 'the largest theft of labor in human history'
- OpenAI head of ChatGPT wrote products 'are largely substitutive, period'
- Internal emails and sworn testimony undercut OpenAI and Microsoft's fair-use defense
- These admissions weaken legal defense in ongoing litigation
- Implication: AI training economics may require licensing or new legal framework
The hook
Internal emails from Microsoft and OpenAI executives call their own training practices 'astonishing theft' — undercutting their fair-use defense in court.
Internal emails and sworn testimony undercut OpenAI and Microsoft's fair use defense. A Microsoft director described the practice as the "largest theft of labor in human history," while OpenAI's head of ChatGPT wrote that the products "are largely substitutive, period."