Not long ago, the police in Baotou, Inner Mongolia, reported that fraudsters used AI face changing and onomatopoeia technology to pretend to be acquaintances to commit fraud, and Mr. Guo, the legal representative of a company in Fuzhou City, was defrauded of 4.3 million yuan in 10 minutes. In another case, in the Anqing Economic Development Zone in Anhui Province, fraudsters swindled 2.45 million yuan from victims using just a 9-second AI face-swapping video. These two typical cases have aroused widespread attention and discussion.
At present, with the rapid development of artificial intelligence technology (AI), new crimes such as AI face swapping scams are also on the rise. Using deep learning and generative models, scammers can digitally fuse the faces and voices of others to create photorealistic composite images. Deepfakes of face-swapping videos and sound clips often cause victims to let their guard down, leading to deception and financial losses.
The good news is that for AI digital technology crimes, relevant legislative research at the national level is continuing to actively advance.
This year, relevant management measures have been issued one after another
Since the beginning of this year, management measures for the application of AI technologies such as deep synthesis and automatic generation have been frequently released.
In January, the Cyberspace Administration of China, the Ministry of Industry and Information Technology, and the Ministry of Public Security jointly issued the Provisions on the Administration of Deep Synthesis of Internet Information Services, detailing the main responsibilities of deep synthesis technology supporters, service providers, service users, and application distribution platforms.
In April, the Cyberspace Administration of China (CAC) solicited public comments on the Measures for the Administration of Generative Artificial Intelligence Services (Draft for Comments). The Measures establish a standardized management framework for AI services, requiring enterprises or individuals of generative AI products or services to conduct real-name authentication, information disclosure, declaration security assessment, performance of algorithm filing and modification, and cancellation filing procedures. The gradual implementation of the above provisions and measures should be able to curb the behavior of illegal individuals or organizations to carry out AI fraud to a certain extent.
On June 25, the Legislative Affairs Commission of the Standing Committee of the National People's Congress held a press conference to address the issue of using "AI face changing" technology to synthesize videos and then commit fraud, spokesman Zang Tiewei pointed out: "The act of video synthesis and fraud through 'AI face change' is a fraud using new technology, which is essentially no different from traditional fraud." ”
It should be noted that the anti-telecommunications network fraud law, criminal law and other laws formulated at the national level have provided more adequate support for combating and controlling various types of telecommunications network fraud, and fraud behaviors such as AI face changing, AI onomatopoeia, and AI phishing emails have not escaped the scope of existing laws.
In addition to legislation, collaborative governance is indispensable
At present, with the maturity of AI technologies such as deep learning, AI fraud presents more and more diverse methods, greatly reduced thresholds, increased difficulty in identification, and the cost of crime is also very low. On the Internet, some high-fidelity face-changing paid software sells for only a few hundred to several thousand yuan, and a large number of leaked personal information and data can also be obtained at a lower price. This may lead to the generation of a large amount of infringing content, which in turn poses a huge threat to citizens' portrait rights, reputation rights and privacy rights, which cannot be completely solved by legislation alone, and requires the coordination of corresponding administrative governance measures.
The first is to vigorously strengthen the capacity building of regulatory bodies. Use machine learning, data mining, blockchain and other technologies to strengthen the supervision and review of AI applications, and timely detect and respond to new criminal methods.
The second is to strengthen technological research and innovation. Develop advanced defense tools and technologies to detect, identify, and prevent fraud crimes using high technologies such as AI. For example, develop strong authentication technologies, fraud detection algorithms, and anti-fraud systems.
At the same time, it is also necessary to vigorously strengthen cross-departmental, cross-industry and even cross-border collaboration to build a cybersecurity ecosystem. Establish joint working mechanisms for government departments such as public security, industry and information technology, internet information, and market supervision departments, as well as telecommunications enterprises, communication software providers, banks, and other enterprises, to provide technical solutions, promote industry standards and best practices, and jointly monitor and maintain network security.
What ordinary people can do
More importantly, every resident is the first person responsible for the security of personal information, and it is necessary to strengthen the awareness of protection and prevent information leakage, especially not to easily provide personal biometric information such as faces and fingerprints to others. Here are some suggestions that may help individuals identify AI face-swapping scams.
Unusual requests or communications: If you receive a request or message that pops up and contains a request to perform unusual behavior, such as transferring money, providing personal information, etc., be vigilant and not respond easily.
Anomalies in language and behavior: Although AI face-swapping technology can realistically mimic the language and behavior of the original character, there may still be some anomalies. Such as incoherent statements, strange grammatical errors, attitudes and behaviors that do not correspond to usual, etc. If you feel that the other person's language and behavior are slightly "unusual", be alert immediately.
Careful observation of facial features and movements: The use of AI face swapping technology to imitate the facial features and movements of the original character is currently not 100% perfect in details, so it is necessary to pay attention to abnormal details, such as uncoordinated eye movements, unnatural facial expressions, and mismatched lip shapes. If a similar situation is observed, its authenticity should be doubted.
Carefully verify identity and information: When receiving requests for important decisions or sensitive information, it's important to insist on verifying the identity and information of the other person as soon as possible, such as direct phone or face-to-face communication, rather than relying solely on online communication.
Watch out for platform warnings: Some social media and online platforms have taken steps to identify or warn of AI-generated content generated by face swaps. If you see urgent alerts about a specific account or content on social media or online platforms, stay alert.
At the same time, always be careful not to leak personal information in daily life. For example, some time ago, many vendors on the streets of Shanghai held cards with QR codes printed on them to talk to passers-by, saying that as long as they scanned it, they could give goose dolls for free. Seemingly insignificant requests and cute geese can easily make people lose their vigilance. However, with a single swipe, stealing personal information, implanting malware and viruses, and illegally accessing your phone's photo album, microphone, etc. can follow.
Therefore, everyone should be vigilant and cautious at all times and not arbitrarily believe and execute unusual requests. If you suspect you've encountered an AI face-swapping scam, it's best practice to pause communication, verify their information and identity, consult a professional, and report it to law enforcement in a timely manner.
Governance and prevention of new fraud crimes such as "AI face-changing" still require multi-pronged efforts by all parties such as legislation, governance and public education, and reasonable guidance of technology for good, in order to deal with potential threats and maintain social security and stability.
--------
Cities are born out of agglomeration.
A city's public policies, living environment, and customs shape the background of citizens' lives.
The Paper focuses on public policy, responds to public concerns, and explores urban issues.