Archive — US — techcrunch
Federal judge sides with Meta in lawsuit over training AI models on copyrightedbooks
A federal judge ruled in favor of Meta, concluding that training AI on copyrighted books by 13 authors, including Sarah Silverman, qualifies as 'fair use.' The ruling emphasized that plaintiffs failed to demonstrate market harm. This follows a similar decision for Anthropic, signifying an ongoing shift in legal standing for tech firms in AI training cases, though specific conditions for such defenses remain critical.
Key points
- Federal Judge Vince Chhabria ruled that Meta's training of AI models using copyrighted books is legal under the 'fair use' doctrine.
- The lawsuit was initiated by 13 authors including Sarah Silverman, who claimed that Meta illegally used their works.
- Chhabria's ruling indicated that the plaintiffs failed to present sufficient evidence that Meta's actions harmed their market.
- The ruling is part of a broader trend favoring technology companies over copyright disputes, as seen in a similar case involving Anthropic.
- The decision specifies that this does not universally validate all AI trainings on copyrighted material and emphasizes the importance of case details.
Why it matters
- This case reflects the ongoing legal debates regarding copyright in the age of AI, similar to earlier disputes in music and film sectors.
- Historically, copyright laws have struggled to keep pace with technological advancements, making this a pivotal moment in defining fair use.
- Current trends in AI development suggest that tech companies will continue facing legal challenges regarding intellectual property rights through 2025.
The ruling favors Meta, showing a potential shift in legal interpretations of 'fair use' regarding AI, but emphasizes that each case's unique details will continue to play a decisive role.