Ex-FTC Chair Lina Khan Urges Criminal Prosecution of AI CEOs Citing New Deal Precedents

Former Federal Trade Commission Chair Lina Khan has reignited the fiery debate over artificial intelligence governance, demanding that federal prosecutors and state attorneys general prepare criminal indictments for AI executives whose systems inflict severe real-world harm. Invoking foundational statutory precedents from the 1934 Securities Exchange Act and early 20th-century anti-monopoly frameworks, Khan insisted that existing American jurisprudence already provides all the legal tools necessary to hold tech leaders personally accountable.

Speaking amidst mounting friction over frontier model safety and corporate concentration, Khan dismissed claims from Silicon Valley that novel technological complexity requires an entirely new legislative apparatus. She argued that when technology leaders knowingly push unverified, autonomous agentic systems into public environments—resulting in unauthorized cyber attacks, widespread financial fraud, or infrastructural compromise—executives should face individual criminal culpability rather than nominal corporate fines.

Crucially, Khan directed intense scrutiny toward the sprawling web of circular investments knitting together Nvidia, Microsoft, OpenAI, and Anthropic. In her assessment, complex joint ventures and cloud compute equity swaps function effectively as synthetic trusts, shielding executives from market discipline while clouding ultimate legal responsibility. By treating rogue agentic outputs as unavoidable probabilistic glitches rather than product defects, major developers have sought to evade the strict liability traditionally imposed on dangerous industrial products.

Khan's rhetoric arrives as several state attorneys general begin exploring criminal negligence theories against software firms whose autonomous tools breached external networks. For enterprise teams engineering autonomous workflows, the legal climate is shifting rapidly. The era of "move fast and break things" without personal consequence is colliding with aggressive regulatory scrutiny, making robust deterministic guardrails, zero-trust architectures, and strict human authorization protocols non-negotiable prerequisites for enterprise deployment.

• The Register - Ex-FTC Boss Khan Urges Uncle Sam to Break Out Handcuffs for AI CEOs
• The Independent - Punish Tech Firms for Rogue AI Incidents, Ex-Regulator Urges
• ANI News - Existing US Laws Can Hold AI Firms, CEOs Liable for Dangerous Products: Former FTC Chair