AI groups could team up on security without facing certain legal trouble.
This bill would let artificial intelligence (AI) companies and other groups work together on security risks without facing certain legal challenges. Currently, teaming up on things like sharing information or delaying AI releases could lead to legal trouble under competition rules. This bill would create clear rules for when they can work together on AI safety without breaking those competition rules.
Today, groups working on artificial intelligence (AI) security must follow existing antitrust laws, which can make it risky to work together on things like sharing threat information or coordinating delays in AI development. If this bill becomes law, these non-federal groups would get a limited exception. This would let them work together on specific AI security purposes without breaking antitrust laws, as long as they meet certain conditions, such as telling the Department of Justice about coordinated delays.
S 5105 · 119th Congress · July 23, 2026 · AI Summary by gemini-2.5-flash · 6/10
Sign in to see your representatives' phone numbers