Bill > HR6881

This bill directs the Federal Trade Commission (FTC) to establish standards for making publicly available information about the training data and algorithms used in artificial intelligence (AI) foundation models. These foundation models are large AI models that are trained on broad data and can be applied across a wide range of contexts. The bill requires covered entities (providers of foundation models with significant monthly output or users) to submit certain information about their models to the FTC, which will then make some of that information publicly available. The required information includes details on the training data, data governance procedures, model capabilities and limitations, and performance on various benchmarks. The FTC is also required to consult with other government agencies and stakeholders in developing the standards and guidance, and to regularly update the standards. The bill provides the FTC with enforcement authority and authorizes appropriations to carry out its provisions.