The Plot Twist …Meta and Anthropic's walk into a library…
The Plot Twist …Meta and Anthropic's walk into a library…
Two federal judges delivered back-to-back AI copyright rulings last week—and while Meta and Anthropic are celebrating, the real story reveals how fragile these "wins" actually are.
Judge William Alsup ruled that using millions of purchased and scanned books to train Anthropic's Claude model is "exceedingly transformative" under fair use. But here's the catch: he was deeply troubled by Anthropic's pirated book collection and ruled that storing those stolen works was not fair use. Anthropic still faces trial over that digital library. 📚
The very next day, Judge Vince Chhabria found that Meta's use of books to train LLaMA fell under fair use, but was far less concerned about piracy. Yet Chhabria made something crystal clear: his ruling only protected these specific plaintiffs. He explicitly noted that future authors could still succeed against Meta using a "market dilution theory"—arguing that AI outputs unfairly compete with original works even if they're not identical copies.
The strategic takeaway? These aren't precedential victories—they're narrow escapes with roadmaps for future attacks. Legal teams building AI programs need to audit data sources and prepare for market harm arguments that go beyond simple copying. The courts just taught plaintiffs exactly how to win next time. ⚡
Comment, connect and follow for more commentary on product counseling and emerging technologies. 👇
