Bruce Willis has not purchased the rights to his face for use in films or television commercials. His agent decided after a series of popular reports stated that he had agreed with a deepfake corporation to duplicate his likeness in upcoming films using synthetic intelligence.
As more studios decide to use AI to replicate famous person likenesses (and sounds) in movies and advertisements, the legalities are likely to become more complicated.
The confusion regarding the Die Hard celebrity appears to have stemmed from this tv commercial final year:
The actor has denied entering into any long-term contract with Deepcake, a Russian company that claims to make virtual replicas of celebrities. Deepcake demonstrated that I had been under the wrong impression. “Deepcake does not own any rights to the virtual dual that the corporation creates,” a company spokeswoman said.
Furthermore, Willis has been diagnosed with aphasia, a speech disorder. He announced his retirement from the stage in March. In theory, AI can correct this by training itself on older recordings of his voice and synthesising that into new content. It has been done for other notable actors, but there is no plan to try it for Willis.
According to Fairness UK, a union for British actors and voice artists, artists are confused by contracts for doing AI-linked voice paintings. Can a player as big as Willis lose not only his face, but also his voice — at the same time as ill?
The core issue is that, technically, faces and voices will be replicated indefinitely by AI. “Contract provisions [for AI work] consistently request that performers sign away their rights in perpetuity,” says Liam Budd, coverage officer for Fairness. In April, the union launched a campaign to strengthen performers’ rights amid a “massive expansion of AI throughout the leisure industry,” according to Budd, who added that 93% of voice actors saw AI as a threat to their jobs.
To make matters worse, performers are frequently asked to sign non-disclosure agreements for AI work, and their contract terms are extremely complicated. Voice actors are also given one-time payments for their efforts, even if their voices are likely to be used frequently in the future.
Much of the progress in AI artwork has come from video games and television commercials. In India, for example, Bollywood superstar Shah Rukh Khan had his voice and face replicated over 40,000 times in a Cadbury’s advertisement to allow him to say a variety of native companies.
That presents a tantalising prospect for big-name actors and studios: AI can help spread their likenesses and “manufacturers” everywhere, allowing them to do more work when they don’t have time.
“We now have instances where well-known voice actors and large motivational audio system want to scale themselves to different areas,” says Alex Serdiuk, the Leader Government of Respeecher, a Ukrainian AI company that de-aged the voices of Mark Hamill and James Earl Jones for recent releases of the Megastar Wars franchise. The generation of speakers can also translate an individual’s speech into other languages. “High-profile individuals are developing their very own logo and IP, and by utilising the generation, they may be able to build up the target market they duvet,” Serdiuk added.