Bipartisan Senate Bill Seeks to Criminalize AI-Generated Digital Impersonation Fraud

A bipartisan Senate bill would criminalize the use of AI-generated deepfakes for financial fraud and identity impersonation, while directing the National Institute of Standards and Technology to lead a federal working group on combating voice and video deepfake technologies.

The AI Fraud Accountability Act, designated S.3982, was introduced by Senators Lisa Blunt Rochester and Tim Sheehy. The legislation targets individuals and organizations that deploy AI-generated voice clones, synthetic video, or other digital replicas to impersonate real people for fraudulent purposes, including financial scams, extortion, and identity theft.

Under the bill, the Federal Trade Commission would receive explicit enforcement authority over AI-powered impersonation schemes. The FTC has already pursued enforcement actions involving AI-generated robocalls and voice cloning scams, but sponsors say existing legal frameworks lack clear criminal penalties specifically tailored to AI-enabled digital impersonation. S.3982 would establish federal criminal penalties for using synthetic media to defraud, extort, or otherwise cause financial harm to individuals or organizations.

The NIST-led working group would bring together federal agencies, industry representatives, and academic researchers to develop technical standards and detection benchmarks for AI-generated audio and video. The group would be tasked with publishing accuracy benchmarks for deepfake detection tools and recommending minimum security standards for identity verification systems exposed to synthetic media attacks.

The bill follows an earlier legislative effort in which Senators Klobuchar and Capito introduced the AI Scam Prevention Act, which targeted AI-enabled impersonation but lacked criminal penalties and a dedicated NIST research mandate. S.3982 expands on that framework with stricter enforcement provisions and a broader definition of covered synthetic content that includes AI-generated audio, video, and image-based impersonation.

Voice cloning and video deepfakes pose direct challenges to biometric authentication systems that rely on voice verification or video-based identity proofing without robust liveness detection safeguards. The Deepfake Summit launched earlier this year to coordinate industry and academic research on countermeasures, while startup isVerified emerged from stealth with voice deepfake detection apps designed to identify AI-generated audio in real time during phone calls and video conferences.

Sources: Congress.gov, Sen. Blunt Rochester

–

By the ID Tech Editorial Team