Celebrity deepfakes springing up on the web, and even in advertising, are raising concerns. The technology is advancing in sophistication and commercial interest. Apple was just granted rights by the U.S. Patent Office to “face image generation with pose and expression control” from reference images. This month, video of President Biden was manipulated into a performance of the viral children’s tune “Baby Shark,” while a digital doppelganger for Elon Musk hawked investment opportunities for real estate startup reAlpha Tech. Tom Cruise, Leonardo DiCaprio and Bruce Willis are also among those artificially misappropriated for promotional use without permission.

Some see this as an opportunity that will allow stars and other public figures to maximize time by deploying digital proxies and extending careers beyond a natural lifespan, an approach by which Apple may be seeking to capitalize, as detailed by Extreme Tech. But there are also serious drawbacks, including unauthorized use by those who would have well-known personalities saying and doing things they would not or did not do.

“Some of the ads are broad parodies, and the meshing of the digital to the analog in the best of cases might not fool an alert viewer. Even so, the growing adoption of deepfake software could eventually shape the industry in profound ways while creating new legal and ethical questions,” writes The Wall Street Journal.

In an ideal context, “authorized deepfakes could allow marketers to feature huge stars in ads without requiring them to actually appear on-set or before cameras, bringing down costs and opening new creative possibilities.”

“We’re having a hard enough time with fake information. Now we have deepfakes, which look ever more convincing,” Carnegie Mellon professor of digital media and marketing Ari Lightman told WSJ. While some states including California, Texas and Virginia have outlawed some deepfake usage, some say a comprehensive federal approach is needed.

“Last year, the U.S. National Defense Authorization Act instructed the Department of Homeland Security to produce annual reports on threats posed by the technology,” WSJ writes.

While alert observers would be unlikely to assume the deepfake video of Joe Biden singing “Baby Shark” was anything other than a parody, not everyone is alert, and clips that start as jokes could gain a “nefarious life online,” a proposition that The Guardian calls “deeply worrying.”

Entertainment interests are particularly concerned about the use of deepfakes for unauthorized marketing. While some celebrities have successfully sued for adverting misuse under right of publicity laws, Wired suggests that anyone less than a mega star with the clout to preserve publicity rights or pockets deep enough to fund litigation may be the most vulnerable where deepfakes are concerned.

“There is a real risk that new actors would be especially vulnerable to signing away their publicity rights as a condition of their first contracts,” Queen Mary University of London intellectual property law professor Johanna Gibson told Wired, noting “this power imbalance could be exploited by studios keen both to commercialize image and character” and expose actors to a contract that would legally allow them to be deepfaked, including into content they may find objectionable.

SAG-AFTRA in the U.S. along with “actors unions across the world” have been “fighting against contracts that exploit their members’ naivety about AI,” Wired writes.

Related:
Bruce Willis, AI and the Problem with Deepfakes, Artificial Lawyer, 10/21/22
Bruce Willis Denies Selling Rights to His Face, BBC, 10/2/22
Deepfake Studio Used 34,000 Bruce Willis Images to Create the Actor’s Digital Twin, Variety, 10/3/22
Keanu Reeves Is the Latest Hollywood A-Lister to Get the Deepfake Treatment, CinemaBlend, 10/14/22
BBC Radio Hosts Use Deepfake AI to Swap Voices, Voicebot.ai, 10/26/22