How concerned are you about deepfakes? Here is why you should be updated on them.
A few days ago, a video featuring Bruce Willis in a commercial made headlines, not because the actor retired from acting but because his image was a deepfake. After the video emerged online, media outlets reported that Willis had sold his image to a deepfake company.
However, a few days later, a representative for the actor refuted the story. As advanced as the concept of a deepfake may be, it can be incredibly harmful, especially for spreading disinformation.
A deepfake refers to realistic images or videos of a person (usually a celebrity or politician) created through AI and technology. The celebrity or politician is replaced with their likeness and is manipulated into saying or doing different things.
In the above-mentioned incident, the video shows Willis dressed in a suit and tied to a time bomb, which echoes scenes from his past action movies. Willis had reportedly sold his face to a Russian deepfake company called Deepcake. The company’s website says it provides clients with hassle-free, effective, and disruptive content, adding, “we can hyper-personalize your brand’s message and jump on the arena of performance marketing with A-list celebrities.” According to Variety, the company used 34,000 images of his face to create his “digital twin” for the ad. Willis was seemingly pleased with the outcome. “For me, it is a great opportunity to go back in time,” he said. “With the advent of modern technology, even when I was on another continent, I was able to communicate, work, and participate in the filming.”
Back in March, the action star announced he would be retiring from acting after 44 years in the business. Willis was diagnosed with Aphasia, a language disorder affecting a person’s communication ability. Willis has long established himself as a hero figure with movies such as the Die Hard series, Unbreakable, The Sixth Sense, The Expendables, and Looper.
Daily Mail initially reported the deepfake story, which other media picked up. The Telegraph reported that Willis was the first Hollywood star to sell his rights to create a “digital twin” for onscreen use. However, the actor’s spokesperson later told the BBC that Willis had “no partnership or agreement” with the company.
Computer-Generated Entertainment or Lies?
Replicating a performer’s image via CGI or hologram to fill a gap is one thing, but replacing the performance entirely is a different issue.
For example, Star Wars producers used Carrie Fisher’s old footage when she passed away before completing the latest movie in the series, The Rise of Skywalker. Another Star Wars actor, Mark Hamill, was de-aged digitally to appear as a young Luke Skywalker in The Mandalorian series. Even James Earl Jones's voice was de-aged for the latest Obi-Wan series.
Earlier this year, the performing arts workers union in the UK launched a new campaign called “Stop AI Stealing the Show.” The union accused AI of replacing skilled professional performers with digital replicas. The actress, Talulah Riley, supported the cause, telling the BBC, “As a performer, it is vital that my voice and my image are my own, no matter how easily and cheaply those things can be digitally replicated. I believe that performers must be rewarded fairly for the content we create.”
The problem with deepfake is not about creating a digital twin for the mere purpose of entertainment. Deepfake can be dangerous in the entertainment industry and other professions, mainly when used to spread disinformation online.
On March 2022, the Daily Beast first reported a story where Ukrainian President Volodymyr Zelensky supposedly shared a video telling his soldiers to lay down their weapons. The footage was debunked, mainly because it was described as a “laughably bad” attempt at deepfake. Zelensky also released his video, telling viewers the video online was fake. Facebook even took down the video for violating its policy on manipulated media.
“As we start seeing more and more cheap fakes, deepfakes flood the zone, it’s going to desensitize people and allow bad actors to allege ‘nothing is real on the ground, you can’t trust anything,” said Mounir Ibrahim. Ibrahim is the vice president of impact at Truepic, a photo and video verification platform. He says, “Anybody, in theory, could begin putting out either crappy deceptive media or really sophisticated deceptive media.”
The difference between deepfake and real videos
While the AI-generated videos may look real, Business Insider says there are giveaways to separate fact from deepfake.
According to Business Insider, deepfakes are not as realistic as the viewer may think. One of the clues is that the subject supposedly never blinks or blinks unnaturally. Another clue is that the subject’s hair, skin, or even face may look blurrier or softer than usual. Viewers should also take note of the artificial lighting, meaning the lighting on the subject does not match the surroundings. Aside from lighting, the audio frequently won’t match the subject speaking. This typically occurs if separate audio is created for the deepfake video.
As technology and AI advance, more opportunities for deepfake videos will appear onscreen. Hopefully, more social media platforms will be able to spot the fakes and take them down before causing further harm (whether intentionally or not.) A computer can’t replicate the authenticity of a factual statement or a physical performance.
Where do you see the future with deepfakes? We would love to know your views on it through comments.
Share This Post On
0 comments
Leave a comment
You need to login to leave a comment. Log-in
