For the 2022 Disney+ series Obi-Wan Kenobi, Darth Vader’s voice was not performed by James Earl Jones, the actor who had voiced the character since 1977. Jones, then 91, agreed to let the Ukrainian speech-synthesis startup Respeecher recreate his voice using AI trained on his archival recordings, and he signed off on the use of those recordings before his death in 2024. Respeecher’s clone, blended into the production with the sound team at Skywalker Sound, let Vader sound as he did in the original trilogy rather than as a 91-year-old man.
The backstory was extraordinary. Respeecher, based in Kyiv, completed work on the project during the Russian invasion of Ukraine in 2022, with engineers continuing under wartime conditions. The company has emphasized that the work was done with consent and with the approval of Jones’s family, positioning it as an example of ethical voice cloning rather than impersonation.
The episode became a touchstone in debates over synthetic performers. It paired a beloved actor’s blessing and explicit licensing with a technology that, used without consent, could just as easily impersonate anyone. The same capability that respectfully preserved Vader’s voice is the capability that fuels fraud and unauthorized deepfakes.
Why business readers should care: the Vader case set an early template for licensing a performer’s voice for AI synthesis - consent, archival source material, and family sign-off - that the entertainment industry has since wrestled into contracts and union rules. It marks the line between sanctioned voice recreation and the unauthorized cloning the same tools enable.