U.S. authority declares AI-generated cloned voice robocalls as unlawful

The U.S. Federal Communications Commission (FCC) announced on Thursday that AI-generated voice calls are considered illegal under existing laws. This ruling follows a deceptive robocall that mimicked President Joe Biden's voice, deterring citizens from voting for him during New Hampshire's Democratic primary election.

According to FCC Chair Jessica Rosenworcel, this definitive ruling allows state attorneys general to intensify their efforts against entities deploying robocalls.

Rosenworcel emphasized, "Malicious entities employ AI-generated voices in unsolicited robocalls for various nefarious activities like deceiving susceptible family members, impersonating famous individuals, and circulating false voter information. The individuals responsible for these fraudulent activities are hereby warned."

Prior to this, state attorneys generals only had the power to influence the results of an unwanted robocall generated by AI. However, this new mandate makes it unlawful to use AI in generating voices for robocalls.

Recently, New Hampshire Attorney General John Formella traced a fraudulent Biden robocall to a Texas-based corporation, Life Corp, which is supervised by Walter Monk. Formella revealed that a cease-and-desist order has been sent to the company and a criminal investigation is in progress.

Democratic FCC Commissioner Geoffrey Starks highlighted the challenges introduced by generative AI. "AI-based voice impersonations have added a novel aspect to voter suppression tactics, inducing more credible fake robocalls during election season," Starks mentioned.

"Voice cloning," as termed by the FCC, may persuade the recipient that someone they trust or care about, like a family member, wishes them to take an action they might not otherwise consider.

In 2023, the FCC penalized conservative activists with a hefty fine of $5.1 million for initiating over 1,100 unlawful robocalls before the 2020 U.S. election.

These violating calls propagated misinformation, saying that people who voted by mail would have their "personal details included in a public database. This information could then supposedly be accessed by police departments to resolve old warrants, and by credit card corporations to recover outstanding payments."