AI Fraud Accountability Act

Key Points

• The bill makes it a federal crime to use AI-generated deepfakes—fake voices or videos of real or imaginary people—to commit fraud. Offenders face up to 3 years in prison and fines, plus forfeiture of profits and equipment used in the crime.
• The Federal Trade Commission gains enforcement power to treat AI-powered impersonation fraud as an unfair or deceptive practice, allowing the agency to investigate and pursue civil penalties against violators.
• A working group led by NIST will bring together government agencies, law enforcement, tech companies, and AI experts to develop best practices for detecting and preventing digital impersonation fraud, publishing annual reports.
• The FTC must identify the top 10 foreign countries where AI fraud targeting Americans originates and work with those nations to cooperate on enforcement, with annual reports to Congress on progress.
• The bill explicitly asserts extraterritorial jurisdiction, meaning U.S. authorities can pursue AI fraud cases even when the perpetrators operate from outside the country.
• A savings clause protects First Amendment rights, ensuring the law cannot be used to restrict parody, satire, journalism, or other protected speech—only deepfakes made with intent to steal money or valuables.