Definition
A US Senate discussion draft bill (introduced June 2026 by Sen. Mark Warner) that would require providers of AI agents — software that acts autonomously on a user's behalf on online platforms — to register with the Federal Trade Commission and meet defined security, privacy, and identity standards. The bill would mandate that each AI agent be traceable to the human owner who authorised it, and that users can immediately revoke an agent's access at any time. It would also prohibit agents from repurposing personal data they collect while acting for users for advertising or other commercial purposes.
Why it matters
This would be the first US federal law specifically governing AI agents, creating a formal accountability chain from autonomous software back to the human responsible for it — a foundational requirement for managing liability as agents become standard enterprise tools.