A U.S. senator queried four AI voice-cloning firms—ElevenLabs, LOVO, Speechify, and VEED—seeking details on safeguards against scam use, consent verification for cloning, detection of impersonation of public figures or minors, and whether audio carries provenance watermarks.

The piece argues that detecting misuse after the fact isn’t enough; it urges forensic digital provenance—caller IDs, routing records, device data, and financial traces—to support legal action.

The next moves hinge on how the companies respond and on new FBI data from 2026 to determine whether voluntary safeguards work and how lawmakers should proceed.

Hassan highlighted the FBI IC3 report showing AI-driven scams caused nearly $893 million in losses across more than 22,000 complaints in 2025, underscoring the urgency.

Consumer Reports’ 2025 findings, which found that four of six voice-cloning products allowed cloning without consent, illustrate safeguards gaps and spur regulatory scrutiny.

The article maps a policy path, including the AI Fraud Accountability Act that would criminalize digital impersonation fraud, with potential prison time, and notes Hassan’s deadline for company replies to shape forthcoming legislation.

ElevenLabs said it blocks celebrity/public-figure cloning and uses a mix of automated and human review; other firms had not publicly responded at the time.

The letters request specifics on how misuse is monitored, consent is verified, impersonations are detected, and how watermarking and provenance tracking could assist forensic and prosecutorial work.

A June 2025 New Hampshire grandparent scam, where an AI-cloned voice was used to defraud a victim, serves as a concrete example of real-world harm from tech gaps.