AI Companies Bypass Protections for Training Data
Gael Varoquaux cites lawsuits and Anthropic projects as examples of AI data practices.
TLDR
Gael Varoquaux posted that frontier AI has broken third-party protections to access articles or books. He linked the New York Times lawsuit against Microsoft and OpenAI filed in December 2023 alleging unauthorized use of copyrighted material for training and Anthropic's Project Panama, a book scanning operation started in early 2024 to build training datasets. A linked report states Anthropic updated its Claude crawler documentation to separate bots for training, search indexing, and user requests with different robots.txt visibility options.
Combined views
381
1 Source, first seen 24d ago