The Reliable Artificial Intelligence Research Act of 2025 directs the Department of Homeland Security to launch at least two prize competitions designed to improve artificial intelligence safety and reliability. The first competition will focus on AI interpretability—making it easier for people to understand how AI systems make decisions—while the second will target adversarial robustness, or an AI system's ability to resist attacks and maintain accuracy even when faced with malicious inputs. The bill requires the competitions to begin within nine months and involve collaboration with multiple federal agencies and private sector experts, with the Secretary of Homeland Security able to contract with organizations to help design and run them. Within six months of the first competition's conclusion, DHS must report to Congress on findings and recommendations for advancing these research areas. The legislation authorizes $10 million in funding spread across fiscal years 2026 through 2030 to support these competitions.
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.