A familiar voice can make a song moving—or make a fraudulent phone call frightening. Voice cloning uses generated audio to imitate a particular person’s voice. Its consequences depend not just on how convincing it sounds, but on who authorized its use and what listeners are led to believe. section 4.9 section 5.3
The same capability appears in sharply different settings:
- With the artist’s support, an AI-generated version of Randy Travis’s voice was used for the 2024 song “Where That Came From.” Travis had lost the ability to sing after a stroke. The release also raised questions about control over an artist’s work and legacy. section 4.9
- Without the artists’ consent, the 2023 viral track “Heart on My Sleeve” mimicked Drake and The Weeknd. It drew millions of plays before streaming platforms removed it. section 4.9
- In a 2023 kidnapping hoax, scammers played what sounded like an Arizona girl’s panicked voice to her mother. They had cloned the voice from a short audio clip; no ransom was paid in that case. section 5.3
Cloned voices have also been used to impersonate relatives in WhatsApp scams. Synthetic speech has appeared in election-related robocalls, ads, and disinformation campaigns. These uses make consent, attribution, compensation, and verification practical concerns—not questions that a convincing voice can settle on its own. section 10.12 section 4.9
