Federal Judge Approves $1.5 Billion Anthropic Settlement Over AI Training Data
A federal judge approved a landmark $1.5 billion settlement between Anthropic and authors, marking the first time a legal price has been placed on AI training data derived from pirated books. The ruling highlights growing scrutiny over how large language models are trained using human-generated content. The settlement covers 482,460 works and is expected to pay approximately $3,000 per work. This decision comes amid rising legal pressure globally for stricter consent standards in data usage, similar to the impact of GDPR on privacy laws years ago. Companies like Reddit have already begun monetizing user content through licensing deals—Reddit reportedly earns about $60 million annually from Google by sharing its comment and post database. Dan Novaes, CEO of Mode Mobile, argues this settlement is a turning point in how human data will be valued, potentially growing into a trillion-dollar market. The shift also affects AI development strategies, with companies increasingly relying on specialized human feedback to improve model accuracy rather than just raw text.
Tickers: $DVLT $META $NET $NVDA $RDDT
Topics: AI settlement approval training data user data rights