In the immense universe of the web, a new form of digital threat is manifesting itself, that of DeepFakes. I had the opportunity to discuss it in my cyber column on M6 and RTL on Julien Courbet's show “It can happen to you”. These deceptive videos, using the image and voice of celebrities, have made a notable appearance in the United States, mainly targeting users of the Medicare and Medicaid health insurance programs.

Journalists from the American media 404 Media shed light on this worrying reality in a detailed article. They revealed a playlist of 1600 deepfake videos on YouTube, totaling more than 195 million views. A real treasure for pirates. These videos hijack the image of personalities like Taylor Swift, Steve Harvey, Ice Cube, Andrew Tate, Oprah Winfrey, and others, such as MrBeast, the leader in videos on YouTube. Some DeepFakes use voices generated by artificial intelligence to reinforce the illusion as I was able to show you in deepfakes used in sextortion.

These deepfakes falsely promise various benefits related to the Medicare and Medicaid programs, tricking victims into disclosing their banking information or making financial transfers. We note the apparent ineffectiveness of YouTube, owned by Google, in detecting and blocking this AI-generated content, despite the outrage of users and celebrities concerned.

It has also been reported that similar scams, using the voices of Cardi B, Tom Hanks, Elon Musk, Rishi Sunak, and even Joe Biden, have been identified on other platforms such as Facebook, TikTok, and Instagram.

A striking example is a deepfake ad by Taylor Swift, viewed more than 300 times, promising illusory financial aid as part of so-called “vacation packages” offered by the state. Other AI clones of celebrities, such as Joe Rogan, Andrew Tate, and Oprah Winfrey, have also been used in these deceptive campaigns.

Deepfake videos often use real clips of celebrities followed by fraudulent messages, creating confusion among viewers. These ads redirect to sites offering fake government aid, a scam previously reported by the U.S. Department of Health and Human Services.

Despite the ease of spotting these scams, as demonstrated by a student who identified more than 16 fraudulent ads, YouTube and Google appear to be unresponsive, continuing to generate advertising revenue from these dubious companies. This laxity raises concerns about Google's ability and willingness to protect its users.

This situation highlights a dark side of generative artificial intelligence: its ability to mass-produce low-quality and misleading content. While AI has the potential to revolutionize many industries, its malicious use in these scams highlights the need for increased vigilance and effective regulation to protect users and maintain the integrity of online platforms.

It is also essential that users stop naively believing everything they see online, 7 days a week, 7 days a year.

read the article