Bill > HR3919

This bill directs the Director of the National Security Agency (NSA) to develop a comprehensive "AI Security Playbook" to protect advanced artificial intelligence (AI) technologies from potential theft or compromise by threat actors. The playbook will identify vulnerabilities in AI data centers and among AI developers, focusing on unique cybersecurity risks specific to AI technologies. It will map out components and information that could be valuable to potential adversaries, develop strategies to detect and prevent cyber threats, and analyze the levels of government involvement needed to secure highly advanced AI systems. The NSA Director is required to engage with AI developers and researchers through interviews, roundtable discussions, facility visits, and collaboration with research centers to gather insights. The bill mandates two reports to congressional intelligence committees - an initial report within 90 days and a final report within 270 days of enactment - with both classified and unclassified versions. The playbook will cover critical AI capabilities across areas like chemical, biological, and cyber domains, and define "technology theft" broadly to include unauthorized acquisition through cyber attacks, insider threats, and other means. The bill emphasizes creating strategies to protect national security without mandating specific regulatory actions, and focuses on preventing potential technological espionage by nation-state and other highly resourced actors.

Military Affairs and Security

https://www.congress.gov/bill/119th-congress/house-bill/3919/all-info