AI Guidelines: Testing, Safety, and Trust in Artificial Intelligence Systems
This Act aims to increase trust in Artificial Intelligence (AI) systems by establishing voluntary guidelines for their testing and evaluation. This means AI systems used in public and private services can be evaluated for safety, privacy protection, and error avoidance according to unified standards. The goal is to support more reliable decisions made by AI systems.
Key points
Establishment of voluntary technical guidelines for internal and external assurance of AI systems to ensure their trustworthiness.
Guidelines will focus on consumer privacy safeguards, data quality, mitigating harm to individuals, and documentation of AI system operations.
Creation of an advisory committee to recommend qualifications and standards for independent entities evaluating AI systems.
Requirement for a study on the market capacity to conduct these assurances, including the availability of qualified experts.
Introduced
Additional Information
Print number: 119_S_2615
Sponsor: Sen. Hickenlooper, John W. [D-CO]
Process start date: 2025-07-31