Beware of the "AI face-changing" scam! You won't be fooled after reading this article!

With the development of AI technology, which makes it difficult to distinguish true from false, it is becoming more difficult to detect online fraud. With the disclosure of a high-value AI fraud case, everyone sneers at such cases. This article introduces the details of AI fraud and tells you how to detect it.

"Traps" are everywhere, and "masks" are hard to tell whether they are real or fake! Have you heard about it? Another outrageous AI face-changing fraud was exposed by the media on May 25! This time, the amount of fraud was as high as 2.45 million, and the entire fraud process took even shorter time, only 9 seconds! It's simply jaw-dropping! Is the safety of our property already threatened everywhere? ! Hurry up and learn about these new scams to avoid becoming the next victim!

Following the previous incident of "4.3 million yuan was defrauded by AI face-changing in 10 minutes" which attracted widespread attention from the society, another AI face-changing fraud was exposed on May 25, which made people feel terrified. According to reports, the amount of fraud this time was as high as 2.45 million yuan, and what's more terrifying is that the whole process took only 9 seconds.

These two incidents once again remind people that with the widespread use of artificial intelligence technology, online fraud methods are also constantly upgrading. People need to be more vigilant and cautious, master and identify the "AI face-changing" trap, so as not to become the next victim.

01 The routine is exactly the same: "Familiar friends + familiar voices"

Recently, the Baotou Public Security Bureau reported a case of telecommunications fraud using artificial intelligence (AI). At noon on April 20, 2023, Mr. Guo, the legal representative of a technology company in Fuzhou, received a WeChat video call from a friend. The other party said that a deposit of 4.3 million yuan was required and asked to transfer the money to a public-to-public account.

Because he trusted his friend too much, and the other party used intelligent AI face-changing and voice-onomatopoeia technology, Mr. Guo transferred 4.3 million yuan to the other party's bank card in two installments within 10 minutes without saying a word. Afterwards, Mr. Guo called his friend and realized that he had been cheated, but it was too late to recover the loss. Subsequently, the police in Fuzhou and Baotou quickly activated the stop payment mechanism and successfully stopped the payment of 3.3684 million yuan, but 931,600 yuan was still transferred.

Coincidentally, according to a report by The Paper on May 25, on May 22, Mr. He from Anqing, Anhui Province, also encountered a similar scam. On that day, Mr. He received a video call from an acquaintance, asking him to help transfer money. But after 9 seconds, the other party quickly hung up the phone, saying that he was in a meeting, and also said, "WeChat and phone calls cannot be used, please add QQ."

Later, the "friend" told Mr. He on QQ that there was a project bidding that needed working capital, and hoped that Mr. He would help advance the payment. " Because we had a video call and the person was an acquaintance," "I didn't think much about it and transferred the money." Based on his trust in the "acquaintance," Mr. He did not hesitate and immediately asked his family to transfer 2.45 million yuan to the account designated by the other party. It was not until he called the other party afterwards that he realized he had been cheated.

Screenshot of WeChat of Mr. He, the victim in Anqing, Anhui (Source: Da Wan)

Fortunately, after discovering that he had been cheated, Mr. He quickly called the police. The special police froze the relevant accounts immediately and rushed to Beijing overnight to arrest three suspects involved in the fraud and recovered the 1.32 million yuan that Mr. He had been cheated of.

Looking back at the two incidents, we can see that when scammers use intelligent AI face-changing and voice-sounding technology, they can imitate the voice and image of familiar friends or family members to make victims feel trustful. This "identity fraud" method is easy to fall into the trap because people are generally more likely to trust people they know.

In both cases, the scammers used a similar approach: they entered the victim’s account through WeChat phone numbers and friend identities, and then used AI technology to disguise themselves as people familiar to the victim , thereby gaining the victim’s trust and guiding them to transfer money. This method makes it difficult for victims to distinguish between the real and the fake, because intelligent AI face-changing and voice-sounding technology can make the scammers’ appearance and voice very similar to people familiar to the victim.

Map: Cultural Industry Review Shi Xinyi

Therefore, in daily life, we should be vigilant, not easily trust strangers on WeChat or other social media, and be careful to confirm the other party's true identity. If you suspect fraud, be sure to call the police and seek help in time. At the same time, the government and relevant departments should also strengthen supervision and technical prevention to combat this kind of telecommunications fraud using intelligent AI face-changing and voice-onomatopoeia technology.

02 Four significant features of AI face-changing technology fraud

AI face-changing technology mainly uses software to transfer a person's facial features, expressions, movements and other information to another person's photo or video to achieve a real face-changing effect.

At present, AI face-changing technology has been widely used in film and television production, digital entertainment, virtual reality and other fields, bringing people a more colorful audio-visual experience. However, this technology also provides opportunities for some criminals to take advantage of it, becoming a weapon for criminal activities such as online fraud and identity theft. Compared with traditional telecommunications fraud, AI face-changing fraud has the following four significant characteristics:

1. Lower cost investment

Traditional telecommunications fraud usually requires a certain amount of manpower and material resources, such as recruiting telephone sales personnel and renting office space, and also requires a certain amount of time and space to plan and commit fraud. In contrast, AI face-changing fraud can be achieved through a mobile phone or a computer and some software, which is low-cost and can be carried out quickly. This also means that fraudsters can obtain benefits more quickly and are difficult to track down.

Map: Cultural Industry Review Shi Xinyi

2. Higher chance of success

Traditional telecommunications fraud usually uses methods such as impersonating public security, judicial organs, bank customer service or prize redemption platforms to intimidate or deceive, and the success rate is relatively low; AI face-changing fraud uses emotional cards to be more realistic and more likely to trap victims. This method often hits the other party's emotional soft spot, such as saying that they are sick and need treatment, or that the company needs support for bidding for major projects, etc., so that the victim will feel sympathy and be deceived. This also leads to a higher success rate for AI face-changing fraud, and victims are more likely to be deceived.

Map: Cultural Industry Review Shi Xinyi

3. More character disguises

In addition to the previously mentioned use of AI face-changing technology to impersonate acquaintances or friends to commit fraud, some criminals use AI technology to disguise themselves as leaders' QQ or WeChat accounts to commit fraud. After gaining trust by adding subordinates or service objects, they ask the other party to help with turnover or transfer money to a designated account, claiming that relatives or family members are in urgent need of money. The victims are in awe of their superiors and dare not question their instructions, making them easily deceived. Recently, the Public Security Bureau of Sandu County, Guizhou Province, successfully cracked a WeChat fraud case impersonating a leader and recovered 190,000 yuan in defrauded funds.

The above is a reference picture of WeChat fraud (Source: Internet)

4. More difficult to prevent

In recent years, the state and local governments have developed some mature prevention measures against traditional telecommunications fraud, especially financial institutions have diversified their means of preventing traditional telecommunications fraud, and social awareness of prevention has also been continuously enhanced. However, there is currently no good solution for AI face-changing fraud. This is because fraudsters can use various social media platforms to commit fraud, and it is difficult to distinguish the authenticity through simple judgment, making it difficult to track and arrest them in a short period of time.

Map: Cultural Industry Review Shi Xinyi

03 Identify and beware of 6 new tricks of AI fraud

In summary, AI face-changing fraud is lower in cost, faster, with more characters, higher probability, and more difficult to prevent than traditional telecommunications fraud, and it also causes greater harm to society. Therefore, we must be highly vigilant against the six new tricks based on AI fraud.

1. Fake audio scam

Using AI technology to generate false voice messages, impersonating relatives, friends, etc. to commit fraud. For example, criminals used AI technology to generate a false voice message from a relative, claiming that the family was in urgent need of an emergency transfer. The victim heard the voice of a relative, which caused panic and anxiety, and he transferred the money without thinking too much.

Image source: Xinhuanet

2. Fake video scam

AI technology is used to generate more realistic fake videos, create more realistic characters or situations, and make victims more easily deceived. For example, in recent years, criminals have replaced a real face with a movie star's face and spread it through social media, inducing users and fans to click on links to download malware or enter personal information.

3. False threat scam

According to Guangming.com, recently, many netizens received the same fake threatening fraud text message. The text message claimed that the other party had been located by law enforcement agencies for credit fraud and required them to pay off all outstanding debts before 14:00 that day, otherwise they would be forcibly "shot" at their door. However, this "low IQ" fraud method is very likely to be AI screening out the most vulnerable people, especially those with low education and little social experience.

Image source: Xinhuanet

4. Fake customer service scam

Using AI technology to generate fake customer service dialog boxes, impersonating banks, payment institutions, etc. to commit fraud. For example, in one fraud case, the scammer pretended to be a bank customer service representative, contacted the victim through WeChat, and lied that the other party's bank account was abnormal, asking the victim to enter the bank card password, resulting in the victim's bank card being stolen.

A scammer pretends to be a JD customer service representative to conduct fraud (Source: Shenzhen Anti-Fraud Center)

5. Phishing Scams

AI technology is used to generate fake websites and emails to trick users into entering personal information or downloading malware. For example, in one phishing incident, hackers used AI technology to generate a fake bank website, and sent text messages to tell the victim to open a bank website and enter their bank card number and password for verification, resulting in the victim's bank card being stolen.

6. Conduct targeted attacks

Hackers can use AI technology to analyze the behavioral characteristics and personal information of target users, conduct precise targeted attacks, and steal sensitive personal information. For example, in a targeted attack, hackers used AI technology to create a fake social media account and sent phishing links through the account, successfully stealing the target user's personal information.

These new fraud methods may be more covert and efficient, posing a greater threat to the public. Therefore, we need to strengthen the supervision and management of AI technology, raise the public's safety awareness, and jointly deal with new fraud methods that may appear in the future.

04 AI face-changing and other fraud prevention guidelines and measures

On May 24, the Internet Society of China issued an announcement stating that with the increasing number of deep synthesis products and services, the use of fake audio and video such as "AI face-changing" and "AI voice-changing" to commit fraud and defamation is common. The general public needs to be vigilant and strengthen prevention.

In this regard, Cultural Industry Review (ID: whcypl) believes that preventing and responding to fraud tricks such as AI face-changing requires simultaneous efforts from multiple dimensions such as individuals, policies, and technology in order to effectively prevent or reduce the occurrence of similar fraud cases.

1. At the personal level, follow the "eight don'ts" and "three don'ts"

For individuals, in addition to staying sober among all kinds of "human skin masks" and avoiding falling into false propaganda, it is more important to maintain a sense of vigilance in the face of ever-changing AI frauds, and to actually "don't listen, don't believe, and don't transfer money."

(1) Protect personal information and follow the “eight don’ts”

First, do not easily provide personal biometric information such as face and fingerprints; second, do not easily disclose your ID card, bank card, verification code and other information; third, do not store ID card and bank card photos directly in your mobile phone; fourth, do not click on unfamiliar links; fifth, do not download unfamiliar software (APP); sixth, do not add strangers as friends casually; seventh, do not scan unknown QR codes casually; eighth, do not enable mobile phone screen sharing with strangers.

Map: Cultural Industry Review Shi Xinyi

(2) Identify false identities and avoid the following three things:

If someone claims to be a "relative or friend," "teacher," or "superior," and contacts you via WeChat, phone, video, or social software such as QQ, and asks you to transfer money with reasons such as "emergency," "private affairs of the leader," or "misfortune," please be alert at the first time. In any case, no matter who the other party is, you must not be afraid of trouble, offending people, or delaying things. Be sure to verify the other party's identity through multiple channels, and never transfer money at will. Please remember, "It is better to take a hundred steps than to take a risky move."

Image source: Xinhuanet

2. At the policy level, we should strengthen the following four aspects:

In December 2022, the Cyberspace Administration of China and three other departments issued the "Regulations on the Management of Deep Synthesis of Internet Information Services", which clearly restricts the generation, replacement, manipulation, synthesis of human voices, and imitation of voices. However, given the current situation and methods of AI face-changing fraud, further prevention and governance are needed.

(1) Strengthen supervision of AI face-changing technology

Further enrich and improve relevant laws and regulations, scientifically standardize the development, use and management measures of AI face-changing technology, clarify the legal scope of use of AI face-changing technology, and severely crack down on illegal activities such as fraud, cyber violence, and privacy infringement using AI face-changing technology.

(2) Strengthen the development of technical preventive measures

Increase the intensity of technical research on AI face-changing technology and the development of preventive measures, improve the security and reliability of AI face-changing technology, develop more intelligent recognition systems to distinguish authenticity, and reduce the living space for such fraudulent activities.

(3) Strengthen cross-departmental collaboration

Public security organs, cybersecurity departments, financial institutions and other departments should strengthen communication and collaboration to form a joint force to jointly maintain the Internet ecology and social harmony and stability. For example, financial institutions should conduct multiple verifications through customer IDs, mobile phone numbers, email addresses and other information to promptly detect and handle abnormal transactions.

(4) Strengthen public safety awareness education

AI-synthesized faces and fake audio and video usually have some characteristics, such as unrealistic images and incoherent voices. Relevant departments should strengthen the publicity of cases of crimes committed using AI face-changing technology, enhance the public's safety awareness and self-protection awareness, and avoid becoming victims of fraud.

Image source: Xinhuanet

In short, AI face-changing fraud is a new type of cybercrime that poses serious social harm. Therefore, in addition to raising individual awareness of prevention, comprehensive measures are needed in terms of technology, policy, law enforcement, education, etc. to strengthen supervision of AI technology and crackdown on criminal behavior, so as to effectively reduce the occurrence of similar incidents and protect the legitimate rights and interests of the public and the security and stability of society.
Author: Xiaojunge Editor: Time Editor: Peninsula

Source public account: Cultural Industry Review (ID: whcypl), China’s top new media for cultural industry.