If your agent commits a crime, who is responsible?

A new analysis from SignalBloom AI examines legal liability when AI agents commit crimes, concluding that current laws are ill-equipped to assign responsibility, and proposes a framework where developers, deployers, and users share accountability based on control and foreseeability. The report highlights that no U.S. statute explicitly addresses AI agent criminal liability, leaving courts to analogize to existing doctrines like strict liability and negligence.