Detect Deepfakesby Resemble AI
Deepfake case study · Audio

Joyce Yost voice deepfake — Nov 2026

Prosecutors in Utah have abandoned plans to use an AI-generated voice clone of murder victim Joyce Yost during a 2027 death penalty re-sentencing hearing

Reported date
Nov 20, 2026
Target
Joyce Yost

The exact incident date was not established. This entry is dated by its source report.

Updated Nov 21, 2026 · 1 min read

Prosecutors in Weber County, Utah, have officially abandoned their attempt to use an AI-generated voice clone of murder victim Joyce Yost to present her testimony in an upcoming death penalty re-sentencing hearing for Douglas Lovell. The decision follows a refusal from the AI software provider to grant the necessary license for the technology.

What happened

To prepare for the 2027 sentencing hearing, the Weber County Attorney's Office sought to recreate the voice of Joyce Yost, who was murdered in 1985. Because no audio recording of Yost's 1985 preliminary hearing testimony exists, prosecutors fed snippets of a police interview recording into an AI tool provided by ElevenLabs. This process allowed them to generate digital audio files that sounded substantially similar to Yost's actual voice, which they intended to use to read her preserved court transcript to the jury.

In 2025, prosecutors requested approval from Judge Michael DiReda to use the AI-generated simulation. The defense team challenged the motion, bringing two expert witnesses to court on August 10, 2026, to argue against the admissibility of the synthetic evidence. Following this hearing, ElevenLabs informed the county that it would not grant a license for the use of its software in this context. Chief Criminal Deputy Attorney Branden Miles subsequently confirmed to the court that the county was unable to secure an alternative AI provider willing to facilitate the request. Consequently, the prosecution has dropped the plan to use the AI simulation, reverting to the use of human proxies to read the victim's testimony.

Sources