ElevenLabs event flags voice-clone issues

ElevenLabs is pitching podcasters on a future where hosts, narrators and ad voices can be cloned, licensed and reused at scale. (elevenlabs.io 1) (elevenlabs.io 2) The company’s podcast tools now promise AI-generated episodes, multilingual localization and cloned voices for production workflows that previously required recording sessions. ElevenLabs says creators can edit, generate and localize podcasts with voice cloning built in. (elevenlabs.io 1) (elevenlabs.io 2) That sales pitch is colliding with a basic disclosure problem: listeners, clients and sponsors may not know when a “host” is synthetic audio rather than a human recording. ElevenLabs’ own materials market podcast creation “without recording a single line yourself.” (elevenlabs.io) Voice cloning works in two tiers on ElevenLabs. Instant Voice Cloning can mimic a voice from short samples, while Professional Voice Cloning trains a dedicated model that the company says can sound “virtually indistinguishable” from the original. (elevenlabs.io) The business model is also getting clearer. ElevenLabs said on April 7 that voice actors had earned a combined $5 million through its Voice Library, with default payouts of about $0.03 per 1,000 characters generated using an approved voice. (elevenlabs.io) That means a podcaster’s voice is no longer just a performance; it can be turned into a reusable asset that other creators, developers or studios license on demand. ElevenLabs says payouts are processed weekly and voices can be tagged, priced and removed by the owner. (elevenlabs.io) The podcast example already exists. Fast Company reported in February that entrepreneur Adam Levy’s AI-generated show “The Epstein Files” had published 97 episodes, uploaded twice daily and reached more than 700,000 downloads in days. (fastcompany.com) Apple Podcasts describes “The Epstein Files” as an “AI-native documentary podcast” built on more than 3 million pages of records and delivered in AI-generated audio. The show says artificial intelligence is used “at every layer of production.” (podcasts.apple.com) ElevenLabs and its allies are also trying to build a cleaner licensing lane for famous voices. Its Iconic Voice Marketplace, launched in November 2025, offers cleared celebrity and estate-backed voices including Michael Caine, Liza Minnelli, Maya Angelou and Mark Twain. (radioworld.com) That does not settle the wider consent fight. A 2024 lawsuit accused ElevenLabs of unauthorized voice cloning, and Senator Maggie Hassan said on April 16 that voice-cloning firms should explain how they verify consent, block impersonation and watermark audio. (legalnewsfeed.com) (forbes.com) The FBI’s Internet Crime Complaint Center said AI-related scams produced $893 million in reported losses in 2025, giving lawmakers a fresh reason to press voice-cloning companies on safeguards. ElevenLabs told Axios, as quoted by Forbes, that it has “an extensive array of safeguards,” including blocks on celebrity and public-figure cloning and review of flagged content. (forbes.com) For podcasters, the immediate question is less whether cloned audio works than whether everyone in the chain knows it is being used. The technology is already being sold as production software, talent marketplace and distribution shortcut at the same time. (elevenlabs.io 1) (elevenlabs.io 2)