Representative Lori Trahan of Massachusetts plans to introduce U.S. House legislation requiring independent safety evaluations of leading AI models and clarifying corporate liability for AI harms