The Guardian reported that ElevenLabs has released a 13-hour audiobook of The Odyssey narrated by an officially licensed AI replica of Caine's voice. The company's ElevenReader listing says the production is an original ElevenProductions audiobook, available for free on ElevenReader, with a full cast performance, original music and sound effects created using ElevenLabs AI audio models. It uses William Cullen Bryant's 19th-century translation of Homer's poem.
Consent is central to the project. The Guardian reported that Caine signed a deal with ElevenLabs in November 2025 to license an AI version of his voice, and ElevenLabs' Iconic Marketplace lists Sir Michael Caine among figures whose AI voices or associated rights can be licensed. In a statement carried by the Guardian, Caine described taking part in ElevenProductions' retelling as a pleasure and framed the audiobook as classical storytelling brought to life through digital technology.
That makes this a cleaner case than the unauthorised voice-cloning disputes that have alarmed actors and musicians. The named performer is alive, the voice replica is marketed as official, and the platform identifies the narration as AI. The hard question begins after that: what kind of performance is being sold when the recognisable voice is licensed, but the line readings are generated?
ElevenLabs presents the answer as a new production model. The Guardian quoted Dustin Blank, who leads partnerships at the company, saying Caine had helped craft the early steps of the technology and that the project should create new kinds of production jobs. The company told the Guardian the audiobook was produced by four producers in six weeks, a far shorter timetable than traditional casting, recording and post-production.
That efficiency is precisely why performer-rights groups remain wary. Equity, the UK performers' union, has campaigned for stronger protections over AI voice and likeness use, arguing that performers need consent, control and compensation when their work is used to train or generate synthetic performances. The Caine audiobook appears to satisfy the first of those concerns for the star voice. It leaves the wider labour question exposed.