DOJ will prosecute crimes committed using AI agents

U.S. Attorney General Todd Blanche said the Justice Department will investigate people associated with artificial intelligence who violate criminal law, according to a statement made at the White House on September 15, 2026.

"As far as AI is concerned, there's a lot of criminal laws. So if anybody associated with AI violates criminal law, we'll investigate that," Blanche said. The attorney general also clarified the department's prosecutorial posture: "We're prosecutors, not regulators," indicating that enforcement would focus on prosecuting existing criminal violations rather than establishing new AI-specific regulations.

Prosecution Under Existing Law

The announcement establishes that the DOJ intends to apply established federal criminal statutes—including wire-fraud and identity-theft laws—to conduct involving AI systems and autonomous agents. The statement does not rely on new legislation or regulatory frameworks, but rather deploys tools already available to federal prosecutors to address crimes enabled or committed by AI actors.

This approach reflects a broader regulatory philosophy in which the Trump administration has signaled resistance to creating AI-specific legal regimes. Instead, Blanche's position emphasizes that existing criminal law is sufficiently broad to address harms arising from AI use—whether fraud, identity theft, wire manipulation, or other violations.

No Specific Cases Announced

The September 15 announcement did not identify any specific defendant, criminal case, court, conviction, sentence, or monetary penalty. The DOJ statement was prospective in nature, establishing a framework for future investigation and prosecution rather than describing enforcement actions already underway.

The timing follows increased public concern about AI-related security incidents and criminal misuse, including deepfake fraud, voice-cloning schemes, and unauthorized access to data systems. By clarifying prosecutorial intent before specific high-profile cases emerge, the Justice Department is signaling to the technology sector that criminal conduct involving AI will face federal enforcement under traditional criminal statutes.

Implications for the AI Agent Economy

For developers and businesses building autonomous agents, the announcement clarifies that the DOJ will not establish a separate regulatory pathway for AI crimes. Instead, companies face existing liability frameworks: if an agent system is deployed to commit wire fraud, identity theft, or other crimes, its operators and potentially its creators face prosecution under conventional federal law.

The statement does not impose new compliance obligations, technical standards, or disclosure requirements on AI developers. Rather, it reinforces that criminal intent and criminal outcomes—regardless of whether they involve AI—remain within federal prosecutorial jurisdiction.