Within weeks of launch, people were cloning Joe Biden's voice. A fake Biden robocall told New Hampshire voters to stay home before the 2024 primary. The voice was generated using ElevenLabs. 4chan users cloned celebrity voices for racist audio. ElevenLabs added verification requirements — after the damage was done. The fundamental problem cannot be solved: a few seconds of anyone's audio is enough to clone their voice. ElevenLabs made this capability available to anyone with a credit card. Once voice identity is unreliable, phone-based verification, voice authentication, and "hearing is believing" are all dead. That bell cannot be un-rung. Three seconds of audio from your social media. That's all it takes to clone your voice. A mother heard her "daughter" crying about a car accident and sent $15,000 in cash. The daughter was fine. The voice was AI. $5 million stolen in 2025. Consumer Reports found most cloning tools have no mechanism to prevent it. Your voice is no longer proof of identity.
What they claim: ElevenLabs states it has safeguards to prevent misuse of its voice cloning technology.
What we found: Within weeks of ElevenLabs' launch, users cloned the voices of Joe Biden, Emma Watson, and other public figures to generate fake audio. A deepfake Biden robocall reached New Hampshire voters before the 2024 primary, using AI-generated voice to discourage voting — the creator used ElevenLabs technology. 4chan users used ElevenLabs to generate racist and antisemitic audio in celebrities' cloned voices. ElevenLabs added verification requirements for voice cloning after the incidents, but the technology had already demonstrated that voice identity is no longer reliable. A voice clone of anyone can be created from seconds of audio.
What they claim: ElevenLabs promotes voice AI for creative and professional use
What we found: The FTC received 36,000+ reports of AI voice clone scams in 2025, with losses exceeding $5 million. Scammers clone voices from 3 seconds of social media audio. One in three people who engage with AI scam calls lose money, averaging $18,000. A Florida woman sent $15,000 after hearing her "daughter" crying about a car accident — the voice was AI-generated. Consumer Reports found most voice cloning products had no technical mechanism to prevent unauthorised cloning. The US Senate sent ElevenLabs a formal letter demanding answers.
What they claim: ElevenLabs privacy policy states it processes voice data to provide its services.
What we found: ElevenLabs' API keys were found hardcoded in the Rabbit R1 device, giving anyone who extracted them access to the complete history of all text-to-speech interactions processed through Rabbit. ElevenLabs is also integrated into Grok Voice, which handles Starlink customer support calls — collecting names, addresses, and account numbers via voice. Every voice interaction processed by ElevenLabs passes through their servers. For a company whose core product is making synthetic voices indistinguishable from real ones, the volume of real human voice data flowing through their infrastructure is a training goldmine — even if the privacy policy says otherwise.