Reliable Artificial Intelligence Research Act Introduced to Congress
The bill, introduced by Representative August Pfluger, mandates the Department of Homeland Security to establish prize competitions for AI safety and explainability.
Representative August Pfluger has introduced H.R. 10594, the Reliable Artificial Intelligence Research Act, aimed at making artificial intelligence systems more understandable and more resilient against attacks.
Purpose of the Act
The Reliable Artificial Intelligence Research Act, numbered H.R. 10594 and introduced by Representative August Pfluger, aims to pave the way for new studies in the field of artificial intelligence. The bill mandates that the Department of Homeland Security organize prize competitions in this area.
Focused Research Areas
The bill focuses primarily on two main research areas. The first is explainability, which examines how artificial intelligence systems make their decisions, and the second is adversarial resilience, which increases the systems' resistance to being deceived.
Department Responsibilities and Timeline
Within 270 days of the enactment of the law, the Secretary of Homeland Security is required to launch two separate competitions regarding explainability for commercial products and resilience in high-risk use cases.
Interagency Cooperation
While organizing these competitions, the Secretary of Homeland Security will consult with the Secretary of Commerce, the National Institute of Standards and Technology, the National Cyber Director, the National Science Foundation, and industry experts.
Budget and Reporting Process
The legislation provides for a total appropriation of $10 million between fiscal years 2027 and 2031. Additionally, a comprehensive report must be submitted to Congress within 180 days following the completion of the first competition.