McAfee antivirus software publishes its ranking of the ten most impersonated personalities on the internet. RMC Conso reveals who these ten celebrities are and explains how to protect yourself against “deepfake”, these hijacked images.
“If you’re watching this video, I’m already dead.” At the end of August, this fake video of Alain Delon created a buzz: if the information inside was absurd, the realism of the voice and the terrifying aspect of using the image of a deceased person to mount a scam was shocking.
Unfortunately, advances in artificial intelligence are making these kinds of videos more and more commonplace. These “deepfake”, false videos which hijack the image of a person to make them say something they did not say, are becoming commonplace.
To alert Internet users and call on them to be vigilant, the antivirus software publisher McAfee is publishing this Tuesday, October 8, its ranking of the ten French-speaking personalities whose image is most frequently misused on the Internet.
Angèle, Timothée Chalamet and Pierre Niney
It is the Belgian singer Angèle who is at the top of the ranking. She owes this first place to the success of her albums and her appearance at the closing ceremony of the Olympic Games, which led to the explosion of her popularity… And scams surrounding her name, her image and her voice .
In second place, we find the star of the Dune saga, Timothée Chalamet, and in third place the famous actor from The Count of Monte Cristo, Pierre Niney.
The full ranking:
- Angela
- Timothée Chalamet
- Pierre Niney
- Lena Mahfouf
- Celine Dion
- Aya Nakamura
- SCH
- Squeezie
- Thomas Pesquet
- Omar Sy
The goal of the usurpers? Make the fans of these celebrities believe that they are really speaking to them, to encourage them to adhere to false information and rebroadcast it, click on a link, provide their personal data, even their bank details, or push them to give money spontaneously.
4 billion threats per day
“This is a real cause for concern for consumers: 63% of French people are more concerned about deepfake this year than last year,” said Vonny Gamot, Europe director at McAfee, contacted by RMC Conso.
“And this is a reflection of a reality: we identify 4 billion threats created by artificial intelligence every day. It was 1.5 billion a year ago.”
More and more stars are raising the alarm themselves when they are victims of such usurpation, like Florent Pagny who, in August, announced that a scam of this type had led to a financial loss of several hundred euros for a fan, convinced to converse and maintain a virtual romantic relationship with the singer.
These “love scams” are one of the first possible traps with “deepfake”: making you believe that your idol has contacted you, that she wants to meet you but that she needs help. She then asks you for money under a false pretext. The most vulnerable and isolated people are likely to fall into this trap.
A second technique used by scammers is to use celebrity to promote a fraudulent site: an alleged online casino for Alain Delon, or even a fake site for the Le Creuset cookware brand for the singer Taylor Swift, whose The image and voice were used without his knowledge last January.
Theft of personal data
Victims who clicked on the fraudulent sites and left their personal data or banking details risk having their money stolen through the use of this data.
Last possible trap: arouse your curiosity with the start of juicy information. To find out more, you must click on a link, which may contain a virus. If you click, the risk is the installation of this virus on your computer, which will allow the hacker to observe your online activity and steal your personal data.
“A simple click can lead to your information being vacuumed up. The golden rule, which I constantly repeat to my parents, is to stop clicking everywhere and anywhere,” recommends Vonny Gamot.
“You may not even realize it right away, because it may be several months before the scammer uses your data, or it may be resold on the darkweb, the black market of the internet.”
According to a recent IFOP survey, only 33% of French people consider themselves capable of detecting a deepfake. However, when in doubt, several clues can put you on the path.
Look for imperfections in the image
The first, and undoubtedly the most obvious, is a speech that is not very coherent, information that is attractive but not very credible, an incentive to make you click, or promises that are a little too good…
Visually, images or voices created by artificial intelligence often contain imperfections: spot a lack of synchronization between the voice and the image, a slightly robotic voice, or even a lack of blinking.
In a photo, observe the person’s body: the parts that are complex to reproduce, such as the hands, are sometimes misshapen. It is not uncommon to have four or six fingers, as in the image below. This photo of a child, published on Facebook, is accompanied by a comment that encourages you to click on a link. It’s a scam.
When it is not celebrities who are being impersonated, scammers use images that encourage compassion, of children or the elderly for example. The objective is exactly the same.
So never click on a website whose reliability you are not sure of, never download software you do not know. Check that the source is serious, and that the information is present on various sources.
If you are vulnerable, protect yourself by installing an anti-virus solution that will also alert you when you encounter an online scam.
The study was conducted by McAfee cyberthreat researchers to determine the number of risky sites and the amount of fraudulent content generated by searching for a celebrity name with commonly used terms. A risk score was calculated for each celebrity.
