Directs the Administrator of the Environmental Protection Agency to establish a consortium that identifies toxic substances and develops safer alternatives through computational toxicology methods based on AI and machine learning.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding legislative act from the U.S. Congress with mandatory language, specific appropriations, and clear enforcement authority delegated to the EPA Administrator.
The document has minimal coverage of AI risk domains, with limited implicit coverage of approximately 2-3 subdomains. The primary focus is on using AI/ML for chemical safety assessment rather than addressing AI-specific risks. There is implicit coverage of system safety (7.3) through validation requirements and potential minimal coverage of governance (6.5) through regulatory framework establishment.
The document primarily governs Scientific Research and Development Services through the establishment of a research consortium. It also has implications for Agriculture, Mining, Construction and Manufacturing sectors as the end-users of safer chemical alternatives, and Public Administration through EPA regulatory decision-making processes.
The document covers multiple AI lifecycle stages with primary focus on Build and Use Model (developing and validating computational toxicology models), Verify and Validate (testing predictability and reliability), and Deploy (making models available for regulatory use). It also addresses Plan and Design through establishing the consortium framework.
The document explicitly mentions AI, machine learning, and computational models for toxicology prediction. It focuses on predictive AI applications rather than generative AI. No specific compute thresholds, foundation models, or open-source model provisions are mentioned.
United States Congress; Senate and House of Representatives of the United States of America
The document is a Congressional bill enacted by the Senate and House of Representatives, making Congress the proposing entity for this governance instrument.
Administrator of the Environmental Protection Agency; Environmental Protection Agency (EPA); Secretary of Health and Human Services; Secretary of Energy
The EPA Administrator is designated as the primary enforcement authority with responsibility to establish and oversee the Consortium, in consultation with other federal agency heads.
Environmental Protection Agency; Administrator of the Environmental Protection Agency
The EPA Administrator has implicit monitoring authority through oversight of the Consortium's activities and the requirement to make results publicly available, though explicit monitoring mechanisms are not detailed.
National Laboratories of the Department of Energy; academic and other research institutions; Supercomputing for Safer Chemicals (SUPERSAFE) Consortium
The Act targets entities that will develop and validate computational toxicology methods using AI, machine learning, and supercomputing. The Consortium members include National Laboratories, academic institutions, and research entities that will develop AI/ML models.
3 subdomains (3 Minimal)