Microsoft's Betrayal: OpenAI Built an Empire on Stolen Words
Microsoft poured over $13 billion into OpenAI while its own executives were documenting what they believed was systematic copyright infringement on an industrial scale.
Newly unsealed court filings reveal that a Microsoft executive privately described OpenAI's data scraping practices as "the largest theft of labor in human history" — even as both companies were simultaneously scraping paywalled New York Times content, building training datasets from it, and warning each other internally that the practice would eventually destroy them in court, according to TechCrunch.
The filings, unredacted for the first time, expose the gap between what the two companies said publicly about responsible AI development and what they said to each other behind closed doors. Microsoft poured over $13 billion into OpenAI while its own executives were documenting what they believed was systematic copyright infringement on an industrial scale.
The timing is precise: Geoffrey Hinton, who built the theoretical foundations on which every major AI model now runs, told Congress this week that lawmakers have roughly one year left to regulate the industry before the technology outpaces any legislative response. The internal Microsoft documents suggest the companies understood this window long before Congress did — and used it.
The New York Times lawsuit, which prompted these disclosures, is still active. What the unredacted filings do is collapse the firewall between Microsoft's public advocacy for AI governance and its private knowledge that the content powering its investments was acquired without consent.
One move for tomorrow: If your business produces original written content, document your copyright ownership formally now. What these filings confirm is that scale does not equal legality — and the litigation cycle on AI-scraped content is just beginning.
