AI face-swapping has defrauded nearly 200 million multinational companies, be wary of these new types of scams
Image source@Visual China
In the new year, in addition to having a good relationship with the God of Wealth, it is also very important to keep the money bag.
The process of technology implementation is often the evolutionary history of pornographic content and fraudulent routines.
Taking advantage of the holiday to have a balance, teach the elders to be vigilant against new AI scams, so as to achieve the "common prosperity" of the family group of "love each other".
Catching a handful of AI bloggers is more deceptive than "fake Jin Dong".
In "Journey to the West", it is said that people are born of people, demons are births, and the AI produced by the code has no parents, but wants to pretend to be flesh and blood, which makes you misunderstand that it is a human being.
Recently, there have been a lot of "Russian beauties" popping up on the video account. They are all fair-skinned and beautiful, fluent in Mandarin but with strange sentence breaks and words, and inaccurate pronunciation of their tones and nasal pronunciation. Of course, considering their foreigner status, it seems natural to have a little accent.
They are enthusiastic and love Chinese culture, and they labeled the video with labels such as "Russian Girls in China" and "Sino-Russian Friendship", and brought "hometown specialties" such as beef tendon sausage, Da Liba, goat milk powder, pickled cucumber, chocolate, and handmade soap to the audience and friends.
In addition to bringing goods, they will also output their views, believing that love has no borders, believing that plain and plain is true, sighing at the prosperity here, and wanting to bring their parents and sister to development.
They may be sisters who have been separated from each other for many years, Yelena and Elena look exactly the same, and the IP is still from Shandong. They gather together as a fire, scattered as a sky full of stars, distributed in all corners of the world, Lena is from Shanghai, Nina is from Anhui, Irene is from Hebei, Katya is from Liaoning, and Alyssa is from Fujian.
These Russian beauties are all AI, and the platform is intimately labeled. We can also distinguish one from the other through their blurred and exaggerated backgrounds, unpicked hair, looping several sets of expressions and movements, and overly down-to-earth lines.
But the middle-aged and elderly people in the comment area can't see it, and really hope that they will stay, and the "Russian beauties" only worry about their own tasks when they reply, completing a quiet soft fraud.
It is human nature to be bewitched by beauty, regardless of age. Xiaohongshu's AI beauty blogger uses fashionable and highly networked characters to achieve the killing of young and old. "Elite man in suit", "White Sox sports student", "Goddess of pure desire", ...... At first glance, I just thought it was a real person who had passed the P picture.
At this time, it is still a way to distinguish whether it is AI or not, because the face is relatively flat and easy to learn, while the hand has a three-dimensional structure, which is more labor-intensive. Needless to say, the number of fingers is exceeded, the knuckles are twisted to keep an eye on the mind, and it is very suspicious not to show your hands unless necessary. Similarly, details such as muscles, limbs, and clothing may have obvious flaws.
The rest of the distinguishing methods are more and more metaphysical and intuitive: the painting style is greasy, the copywriting network suppresses the clouds, there is no breath of life, the light and shadow effects are unnatural, the eyes are dead, the face is too delicate and perfect, there are only static pictures and no videos, it is not easy to determine a single picture, but you can find out if you look at a few more photos, and any posture is an expression......
Xiaohongshu AI blogger @cyberAngel_ is a software engineer by profession, AI painting is a hobby for love to generate electricity, he does not hide the fact of using AI, the title of the post states "AI painting", and the account profile also reads "I am just an emotionless robot".
Image from: Xiaohongshu @cyberAngel_
At the same time, @cyberAngel_ believes that in the future, AI painting will become more and more difficult to distinguish between real and fake, AI generation and AI detection are mutually evolving, and creators are more aware of where AI is not like real people, so they will strive to make this part "like" real people, such as the beautiful girl produced by AI is too perfect, and it can be appropriately blurred to add a "film sense".
Even if the AI beauty blogger is not so real, the plot similar to the "Russian beauty" has appeared, and the platform has clearly prompted that "it is suspected of containing AI creation information", and there are still people in the comment area who leave rainbow farts, and even chase and ask about the link to the clothes, and the question that seems true or false will inevitably give way to beauty is justice in the end.
When the number of fans comes up, the monetization methods of AI beauty bloggers can be very diverse, some are self-reliant, monetize traffic through advertising, etc., some pay for knowledge, sell courses, sell custom models, according to @cyberAngel_'s words, "new media has new media monetization methods, and technology has technical monetization methods".
Live-action model.
The AI "Russian Beauty" and "Beauty Blogger" have some similarities with the "fake Jin Dong" who deceived middle-aged and elderly women of their emotions and money before, but they are more terrifying and more imaginative.
"Fake Jin Dong" has no technical content, the videos and photos related to celebrities are downloaded or packaged and purchased, the affectionate voice is faked with voice-changing software, and it is so rough that it is "fake at a glance", and the AI has a higher rank, and sometimes it is really not so easy to distinguish.
It can only be said that when we see handsome guys and beauties on the Internet, it is best not to default to real people, nor do we have to be too sincere, Internet celebrity companies do not want to spend money on real KOLs, and the audience also consumes batches of beauty and cheap views more quickly, how in line with the communication laws of the Internet.
The AI mimics your face and terrorizes your heart
In addition to casting a wide net through "beauty bloggers" and "Russian beauties", AI can also accurately target.
In other words, AI can not only weave a gentle township, but also transform itself to become a PUA master.
A common type of AI scam is in which scammers fake caller IDs and acquaintance voices to tell middle-aged and elderly people who receive calls that their grandchildren are in trouble, either have committed an accident and need money to eliminate the disaster, or are in danger and need to be rescued with ransom money.
Several such phone scams have been reported abroad, and the plots are similar, with kidnappings, injuries, and drunk driving rear-end collisions...... Similar things also happened in China, especially by taking advantage of the "time and space difference" to fabricate that international students were kidnapped, so as to deceive parents.
The routine itself is not new, it has been rampant for several years, but because of technological development, the requirements for computing power, samples and other aspects have been reduced, the effect is more realistic, and the scam is easier to implement.
ElevenLabs, the leader in AI voice cloning, delivers high-quality audio for just a dollar and a minute of high-quality audio that lets you master 29 languages and multiple tones in an instant, while still retaining your accent, intonation, and rhythm.
While Elevenlabs now repeatedly assures that only you can clone your own voice, and that there is a verification process to prove that the voice belongs to you, this is a remedy, and when Elevenlabs first launched its beta in 2023, clones of celebrities like Taylor Swift were already flying all over the place. What AI Taylor Swift says, I don't have the right to decide.
However, although the other party can fake the caller ID and voice, if the person is sober, hangs up the phone, re-enters the mobile phone number, and takes the initiative to contact the relatives, the lie is often self-defeating.
Compared with voice, video calls may make middle-aged and elderly people who believe in "seeing is believing" doubt their lives.
The old joke of not knowing whether it's a person or a dog on the other side of the screen never goes out of style, and it's not technically difficult to implement. A common operation is that scammers use virtual camera software and AI face swapping to video chat with the other party.
A Shenzhen-based technology company interviewed by CCTV that provides technical support to the public security organs said that when performing real-time face swapping in video chats, whether it is an avatar or a photo of Moments, it only takes about 30 seconds for feature recognition after uploading the image, and then the AI starts modeling, and the modeling is completed for real-time conversion.
One-on-one videos may be wary, but what about a one-to-many "professional team"?
Recently, the Hong Kong branch of a multinational company was defrauded of $25 million (about 180 million yuan) because of AI. The victim was a finance officer who received an email from the "CFO" at the UK headquarters and was invited to participate in a video conference involving "secret deals", at which not only the "CFO" but also several familiar "colleagues" were present.
The scammers downloaded the public video, faked the faces and voices of real people through deepfakes, and applied them to the video conference. In order to avoid revealing the truth, the "CFO" unilaterally ordered the "colleagues" not to communicate with the victim, and the video quickly hung up, and the scammer continued to contact the victim by email and other means.
Police demonstrate how to use Deepfake to fake a multiplayer video conference.
Although there are investigation results in this matter, some netizens suspect that it is just a ghost. It is not known how the risk control measures of multinational companies are, but it is certain that AI face swapping can also extend to many places, from the earliest head swaps to fake celebrities bringing goods in live broadcast rooms.
Thankfully, AI isn't perfect, and it's still easy to tell.
If the video is changed through AI, the original image of the camera consumes computing power when converting several layers, and the sound and picture often have a delay. In addition, you can guide the other person to do some actions, such as opening the mouth, shaking the head sharply, moving your fingers back and forth in front of the face, etc., if the AI changes the face, the "face" may appear deformed and blemishes.
In addition to breaking through the technical bugs, another tried and tested trick is to ask some privacy questions that you know but I don't know, or make up lies and deliberately tricks to see how the other party answers.
Of course, the lack of technology can only solve a temporary dilemma, you might as well strategically despise and tactically pay attention to it, and you can't feel that you will not be deceived, it may just be that the high-end game has not yet appeared.
AI scams are not new, but a new digital divide has emerged
2023 is known as the first year of generative AI, but there is a lag in the adoption of technology, and perhaps most ordinary people can perceive how technology, which seems not to be new, but has become more and more popular in recent years, has affected daily life.
Everyone understands that technology is a double-edged sword, Elevenlabs can make people with language barriers speak, HeyGen allows mildew to lip-sync Chinese without translation, and Miao Duck camera allows people to harvest refined ID photos without leaving home, the bright and dark sides of technology have no long or short edges, but only objectively exist.
However, for many middle-aged and elderly people, smartphones and the Internet are incomprehensible, and now AI is messing with it again.
In the era of AI, text, sound, pictures and videos may be fake, or even in the form of combinations, and the identities forged by scammers are more specific, and the fraud is more targeted and true, making the probability of middle-aged and elderly people being deceived higher.
Magic against magic, technology offensive and defensive technology, is a cat and mouse game, the elderly may not be able to digest quickly, we may start from the essence of fraud, reduce the possibility of the elderly being deceived.
No matter how technology develops, many scams have the same routine: stealing privacy, using fear, greed, and emotional value to make up stories, posing as acquaintances or packaging themselves to gain trust, and trying to get money as the ultimate goal.
After the AI fashion pope came out of the circle, an X (formerly Twitter) big V with nearly 13 million fans sighed: "I thought the Pope's down jacket was real, and I didn't think much about it. There's no way I'm going to survive the future of technology. 」
For middle-aged and elderly people, it is more difficult to adapt to the rule that seeing is not equal to the truth, and having pictures is not necessarily the truth. We can tell them something simple, traditional, but still effective.
One is to be careful to sail a ship for thousands of years, not to easily believe the content of the Internet, not to answer harassing calls, not to click on unfamiliar links, and try not to overexpose personal biological information such as faces, voices, and fingerprints on the Internet.
Especially when receiving suspicious calls and text messages, don't blindly believe in the "one-sided words", no matter who the other party is, when it comes to money, you have multiple eyes, verify the identity of the other party through a variety of ways, if the other party says that something happened to the family, call the family again to confirm whether the situation is true.
The other type is to take the initiative to be prepared, middle-aged and elderly people have their own preferred media, and their families can forward the anti-fraud content of government departments and other sources such as WeChat official accounts to them. If possible, middle-aged and older people can also be shown how to use AI tools such as ChatGPT.
Of course, for the elderly who don't understand smartphones, try to solve the problem at the source, and it is best not to bind bank cards to WeChat and Alipay, and only save a small amount of change.
Successful AI scams are only the final result, but prevention against AI scams can start at any time.
Technology should serve everyone, guide the way to illuminate the beacon of the digital world for those who come after us, and explore the unknown reefs under the water, so that we can collectively move towards a future where technology is not feared but used wisely.