This bill prohibits federal agencies from using or purchasing artificial intelligence models that fail to meet safety standards designed to prevent the generation of child sexual abuse material and non-consensual intimate images. Within 180 days of enactment, federal agencies and their contractors must remove any AI models that don't comply with these standards from their systems, and the General Services Administration must establish ongoing review processes to identify problematic AI models at least every 90 days. The legislation also establishes performance benchmarks through the National Institute of Standards and Technology and creates a private right of action allowing individuals harmed by unsafe AI systems to sue developers and others who circumvent safety controls, with potential damages of up to $50,000 or treble damages for repeated violations. The bill gives vendors an opportunity to modify unsafe AI models to bring them into compliance before federal agencies must remove them, and allows agencies to implement additional safeguards if vendors cannot make the necessary modifications.
Take Action
Your position
Add a comment
to comment on this bill.
Annotate the text
Highlight any passage on the Summary or Full Text tab to attach a note. Annotations appear on the Annotations tab.