US Legislation Demands Transparency on Copyrighted Content in AI Training

Representative Adam Schiff has introduced the Generative AI Copyright Disclosure Act, demanding AI companies reveal copyrighted content used to train generative AI models. Failure to comply could result in fines of at least $5000. The legislation seeks transparency and accountability, requiring formal submission of copyrighted works to the Copyright Office before model release. As AI models scrape diverse internet data, including copyrighted material, disputes have arisen, as seen in the New York Times' lawsuit against OpenAI and Microsoft. The lawsuit alleged unauthorized use of millions of newspaper articles for AI training. The proposed legislation aims to address such concerns, ensuring creators' rights are respected in AI development. If passed, the act will impact both new and existing AI models, compelling companies to disclose their training data sources.