This AI Company Stole My Voice (allegedly)
Key Takeaway
Unregulated AI voice cloning allows platforms to replicate unique speech patterns and accents for commercial use while leaving individuals with no reliable legal referee to prove ownership.
Highlights
A popular default British AI voice named Archer on the Eleven Labs platform sounded identical to podcast host Chris Williamson.
Eleven Labs ran an ENCAPA-TDNN technical similarity test that returned a score of 0.28, which fell well below the 0.7 threshold required to indicate a voice match.
Bette Midler successfully sued the Ford Motor Company in 1988 after an ad agency hired her backup singer to imitate her voice for a commercial.
Tom Waits was awarded $2.4 million after suing Frito-Lay for impersonating his voice in a commercial.
State-level right of publicity laws protect individuals when their voice or likeness is misappropriated for commercial gain.
Timeline
Discovery of the AI Voice Twin
A friend originally sent a video about flight MH370 that immediately drew attention due to the narrator's uncanny resemblance to the host's distinct speech patterns. Pronunciations of words like yours, purple, work, and june carried specific regional markers from the northeast of England that match the host's background. This voice was identified as Archer, a leading British male voice option on Eleven Labs. The ease of democratized voice cloning creates a dangerous environment where an artificial replica can make authentic-sounding endorsements for ideas or products the original speaker opposes. Personal voice functions as the vocal equivalent of a fingerprint and forms the foundation of human trust. When listeners encounter an inauthentic AI clone of a known voice, subconscious alarms trigger a sense of distrust toward the actual person.
- A YouTube video about flight MH370 featured an AI narrator that matched the unique vocal artifacts and regional Middlesbrough accent of podcast host Chris Williamson.
- The voice clone appeared as the default British male voice named Archer on the AI platform Eleven Labs.
- Anyone across the globe could utilize this voice model for a few dollars to make the speaker endorse products or opinions.
- Unlike musical samples or song riffs, public perception has not yet established a standard of disgust around unauthorized voice cloning.
The Eleven Labs Response and the Verification Dead End
Reaching out to Eleven Labs brought a response defending the platform's integrity. The company cited a technical similarity test known as the ENCAPA-TDNN test, which scored the match at 0.28 compared to the 0.7 threshold required for a positive identification. This left the host trapped between trusting his own ears and friends versus accepting a corporate scientific metric that claimed the voices were entirely different. Although the Archer voice was ultimately removed following attention from a popular podcast episode, the underlying uncertainty remained unresolved. Physical property theft involves police reports, and forged signatures involve banks, but voice cloning leaves individuals with unverified customer support emails and ambiguous automated tests.
- Eleven Labs stated their legal and voices teams investigated the original voice actor behind the Archer model and found no fraudulent behavior.
- An ENCAPA-TDNN similarity test returned a score of 0.28 between the Archer voice and the host's voice, where any score below 0.7 indicates distinct identities.
- The disputed voice model was eventually removed from the platform after public complaints and podcast mentions.
- Victims of voice cloning face profound helplessness because current systems lack independent arbiters or referees to verify ownership.
Legal Precedents and the Right of Publicity
Historical context shows that voice imitation lawsuits date back decades, such as Bette Midler suing Ford Motor Company in 1988 and Tom Waits winning a judgment against Frito-Lay. A consultation with a copyright lawyer confirmed that while federal legislation does not explicitly cover voice clones, individual states recognize right of publicity laws. These laws protect against unauthorized commercial exploitation, relying heavily on whether a reasonable person would believe the individual endorsed the product. Documenting public confusion and commercial usage serves as primary evidence for potential litigation. Ultimately, as AI-generated face and voice scams proliferate, the lack of definitive legal frameworks leaves internet users vulnerable to synthetic impersonation without immediate recourse.
- Historical cases like Bette Midler versus Ford Motor Company established that unauthorized voice imitation for commercial products is unlawful.
- Tom Waits secured a $2.4 million judgment against Frito-Lay for commercial voice impersonation.
- Copyright lawyers confirm that most U.S. states enforce right of publicity laws protecting individuals against commercial misappropriation of their likeness.
Description
The robots are coming for our voices… - Get access to every episode 10 hours before YouTube by subscribing for free on Spotify - https://spotify.modernwisdom.com or Apple Podcasts - https://apple.modernwisdom.com Get my free Reading List of 100 life-changing books here - https://chriswillx.com/books/ Try my productivity energy drink Neutonic here - https://neutonic.com/modernwisdom - Get in touch in the comments below or head to... Instagram: https://www.instagram.com/chriswillx Twitter: https://www.twitter.com/chriswillx Email: https://chriswillx.com/contact/
