I guess everyone has heard about this recently: an actor was tricked into a Southeast Asian extortion ring in the name of "auditioning for a movie" and was eventually rescued and returned home. I can't help but sighScamsThe tactics are endless, and if you are not careful, you may fall into a carefully laid trap.
As a platform for AI applications, it suddenly feels necessary to summarize the typical AI scamsA set-up, a warning to myself, and to everyone else.
Feel free to retweet this so that more people are aware of AI scams and alert. 🚨
Routine 1: AI face-swapping video scams
Fraudsters collect information about someone through various channels, such as name, phone number, social accounts, photos, and videos, before using AI face-swapping technology to create an AI digital person who looks like that person.
This tactic often uses high-level, boss image to pressure employees to let their guard down and urge victims to transfer money as soon as possible and so on.
📘 Practical cases
In early 2023, an employee of the Hong Kong branch of Arup, a multinational engineering firm, received an email invitation to a meeting that appeared to be from executives at the company's headquarters. Although he was initially skeptical, he let his guard down when he clicked on the link in the email to enter the meeting and saw that a number of executives, including the company's CFO, were online and discussing a secret acquisition. After the meeting, he followed the instructions of the "leader" and transferred HK$200 million into five designated accounts in 15 installments. It was not until he communicated with the headquarters five days later that he realized he had been cheated.
It turns out that the fraudsters used Deepfake technology to create "avatars" using audio and video messages from company executives and prerecorded the fake high-level meeting, allowing employees to fall into the trap without even realizing it. The case is considered one of the largest AI scams in history.
Set two: AI synthesized voice fraud
Fraudsters obtain the voice material of the target person through harassing telephone recordings and other means. Then they use AI voice synthesis technology to imitate the voice of the target person, communicate with the target person's acquaintances over the phone, and use a variety of urgent or enticing reasons to obtain trust and money.
With the increasing fidelity of tone cloning these days, fake tones are hard to recognize. It is advisable to refuse all requests for remote transfers that are "all voice and no person". 🙅
📘 Practical cases
In 2019, the world's first AI scam took place in the UK: a fraudster used advanced AI voice mimicry software to pose as a company executive and successfully defrauded the managing director of a British energy company out of 220,000 euros.
Recently in China, a company's founder was scammed out of up to 4.3 million yuan in just 10 minutes of using WeChat to video chat with a friend.
Trope 3: AI Generates Fake News
Fraudsters use AI generation technology to generate fake news content based on specific keywords or topics. These contents are very similar to real news in terms of language expression, logical structure, etc., and are highly confusing.
Then, by publishing false financial news, product reviews and other information, people are induced to invest or buy inferior products.
📘 Practical cases
In January this year, Nanjing Sun Mou, Suzhou Li Mou and other people, in order to gain attention and improve property sales and other purposes, fabricated and disseminated "compulsory education reform to shorten the school system", "high school into compulsory education, Nanjing in Xuanwu District pilot" and other rumors, resulting in adverse social impact, disrupting public order. Social impact, disturbing public order. At present, Sun Mou, Li Mou and other seven people have been sentenced to administrative punishment by the police according to law.
🤝 Precautionary advice
When getting news, try to choose the channels released by official media or authoritative organizations. For some shocking secret news and gossip, keep calm and do not spread mindlessly to follow the trend.
Trope 4: AI Scuba Scam
Fraudsters use AI technology to generate a large number of fake accounts and control the activities of these accounts on e-commerce platforms and social media. Then, through the AI writing function, they generate a large number of false reviews and favorable comments to mislead consumers into believing that a certain product or service is very popular.
🤝 Precautionary advice
When purchasing an item, in addition to checking the number of reviews and positive ratings, read the content of the reviews carefully, especially those with descriptions of specific usage experiences and details.
Be wary of items that have seen a spike in the number of reviews in a short period of time and have unusually high positive ratings - there may be a risk of fake reviews.
Trope 5: AI Emotional Fraud
Using AI face-swapping technology and natural language processing technology, it disguises celebrity characters from TV dramas and chats with victims 1v1 through social software.
After gaining the victim's trust, they use emotions as bait and fabricate various reasons to "sell" the victim and ask for money.
🤝 Precautionary advice
When communicating with so-called "celebrities" or "acquaintances", keep your wits about you. "Celebrities are very busy and are unlikely to come over to talk to a stranger, and don't transfer money or provide personal information easily.
Verify the identity through the star's official channels, such as through the official studio, the star's other official accounts, etc.
Trope 6: AI Money Fraud
Fraudsters use AI technology to analyze large amounts of user data to accurately identify potential fraud targets, such as people who frequently focus on financial investments.
Then, customized fraud scripts are generated according to the characteristics and needs of the target group, and false investment information is sent to them by phone and SMS.
Trope 7: AI Phishing Scam
Use AI technology to hack into someone's email & other files, then mimic that person's writing style to send phishing letters to their social friends that are both real and fake.
The emails usually contain malicious links or attachments that entice victims to click or download, and once the user clicks on them, their personal information is stolen.
After reading these cases, have you noticed a pattern: AI scams are not only exploiting technology, but also exploiting human psychological vulnerabilities.
As a result, some have also compared AI fraud to an 'arms race':
Fraudsters are using more advanced AI technology to break through people's psychological defenses and create chaos, while ordinary people need to band together to defend themselves against human weaknesses by reinforcing rules and 'common sense'.