Meta and Anthropic scored key legal wins this week in separate lawsuits over AI training data. Federal judges sided with both companies, backing the idea that using copyrighted books, images, and other creative works to train AI models can qualify as "fair use"—even without permission.
The rulings mark the first real court validation of AI companies’ defense of their training methods. This could reshape ongoing lawsuits and set the tone for how AI firms handle copyrighted material going forward.
The decisions don’t set nationwide precedent—appeals are expected. But legal experts see these wins as a potential turning point in the fight over AI training data. 
On this week’s TechCrunch Equity podcast, hosts Max Zeff and Anthony Ha, joined by Sean O’Kane, discussed the rulings and what’s next for AI companies battling copyright claims.
Anthony Ha noted on the show that more challenges are on the horizon, and broader legal battles will impact how AI interacts with creative industries.
Equity will return next week with more updates on tech and startups.
Equity is TechCrunch’s flagship podcast, produced by Theresa Loconsolo, posting every Wednesday and Friday. Subscribe on Apple Podcasts, Overcast, and Spotify. Follow on X and Threads at @EquityPod.
