Trump EO directs AG to prosecute AI-agent computer fraud
President Donald Trump signed Executive Order 14409 on June 2, 2026, directing the U.S. Attorney General to prioritize federal prosecution of individuals who use AI agents to illegally access or damage computer systems JD Supra. The order explicitly targets those "employing AI agents to unlawfully access data or information that is subsequently used for a criminal or unlawful purpose."
No New Crime—Existing Statutes Weaponized
A critical distinction separates this executive action from criminalizing AI agents outright. The order does not create new criminal liability; instead, it directs enforcement resources toward existing federal law The Agentic Protocol. Section 4 of EO 14409 specifically names three statutes: 18 U.S.C. § 1030 (Computer Fraud and Abuse Act), § 1028 (identity fraud), and § 1343 (wire fraud). The AI agent functions as a tool wielded by a human actor, and that human bears criminal liability under pre-existing law.
The language in the executive order reads: "The Attorney General shall prioritize enforcement of existing federal criminal statutes... against anyone who utilizes AI to illegally access or damage a computer without authorization, or who utilizes AI while engaged in such illegal access to further any other crime." This represents a prosecutorial pivot, not legislative innovation JD Supra.
Scope: Unlawful Access, Not Agent Use Generally
The order's reach is narrower than headline language suggests. It targets unlawful computer access—breaking into systems without authorization—not the use of AI agents in legitimate contexts. Compliance frameworks released in response to the order emphasize that lawful agent deployment for business purposes remains unrestricted Baytec Consulting. The executive order also explicitly disclaims authority to impose mandatory licensing, preclearance, or permitting requirements on AI models, preserving innovation in the agent layer.
Enforcement Authority and Practical Impact
No court has yet charged an individual under this directive. However, the shift from passive enforcement to active prioritization by the Attorney General's office signals increased resource allocation toward cybercrime units tracking AI-agent-assisted intrusions. Penalties remain subject to underlying statutes: CFAA violations carry sentences up to 10 years for repeat offenses causing loss exceeding $5,000. The order affects actors from individual threat groups to organized cybercriminal networks that have begun automating reconnaissance and lateral movement with agentic tools.
For enterprises and developers, the practical implication is clear: deploying agents for any form of unauthorized computer access—credential stuffing, network scanning, data exfiltration—now faces heightened federal scrutiny and prosecution risk.
