4 young people create anti-online fraud tool using AI

The "white hat" knights

Just after school, a group of young students led by Le Nguyen Viet Cuong gathered in a small room and discussed enthusiastically to perfect an AI tool to identify online scam websites. Few people know that although they are 2nd and 3rd year students, they are already key members of the Chongluadao.vn project founded by cyber security expert Ngo Minh Hieu.

Although still very young, the group of technology students including Viet Cuong, Huy Quang, Duc Sang, Vinh Khang have been active members of the anti-fraud project. PHOTO: NVCC

Although quite cautious in his first communication with PV, Le Nguyen Viet Cuong, a second-year student at FPT Swinburne University, still appeared mature and confident. Cuong said: "We have noticed that phishing websites are becoming more and more numerous and sophisticated. Previous prevention solutions, such as relying on community reports or basic elements in the web address, are not enough to deal with these threats. Some of my friends have been victims of phishing websites, specifically websites that impersonate the social network Facebook. The scammers pretend to share a link to a post, but in fact lead to a fake website and ask for a login to get an email and password. After getting the account, they will text their acquaintances to ask for money or continue to spread the malicious link."

"From those pressing realities, my group, under the guidance of Mr. Hieu (cybersecurity expert Ngo Minh Hieu - PV), my group decided to develop a more comprehensive and advanced web analysis tool, integrating artificial intelligence to assess risks based on many factors such as content, source code, screenshots... to produce more accurate results and help protect the community from criminals," said Le Nguyen Viet Cuong.

With the dynamism and creativity of young people, a group of technology students have diligently researched and officially launched an AI analysis tool, integrated on the website Chongluadao.vn and directly experienced by reporters. Recently, when verifying a form of fraud through a service of exchanging scratch cards for cash advertised online, Thanh Nien reporters contacted the Chongluadao.vn group to check the website "doicard.com.vn" and quickly received a response. With characteristics recognized by AI such as providing a service of exchanging cards for cash with suspiciously high discounts, using urgent language while lacking appropriate business registration information, promising an unrealistic 3-second processing time... In particular, this tool also detected that the above website contained extended JavaScript to track user behavior, so it was determined to be potentially malicious and had a very high risk score of 9/10.

Malicious website checking tool uses artificial intelligence to analyze and return results quickly to users. PHOTO: SCREENSHOT

Responding to Thanh Nien reporter , Nguyen Van Huy Quang, a young member of the group, shared: "Fraudulent websites often impersonate famous websites such as social networks, e-commerce, banks, governments... by using URL addresses and content that is almost real. In addition, there are tricks to make users enter sensitive information or transfer money, such as notifying that the account has a problem and asking to re-enter the password, or advertising products at very cheap prices... On the user side, most of them are deceived due to lack of vigilance. With the sophistication of today's fraudulent websites, if you just look at them briefly, it is very difficult to distinguish which websites are safe and which are threats. Moreover, scammers are very good at manipulating victims' psychology so that they easily fall into traps, causing serious damage to many people both physically and mentally. These are the realities that make us very worried and determined to contribute our efforts to the community".

Community cooperation is needed to combat fraud.

Nguyen Vinh Khang, a member of the group, excitedly shared: "Our group was initially guided by Mr. Hieu, because he has a lot of experience in this field. After that, we learned more and experimented to find suitable artificial intelligence technologies and models. Currently, the core features of the application have been completed and can be used effectively with an accuracy of more than 98%. Our group will continue to test and develop to improve the solution, as well as open the source code to attract more contributions from the community. We believe that, with the right direction and effective implementation, this application will bring a lot of value to the community in dealing with today's online threats."

Joint efforts and many AI application projects are needed to effectively combat online fraud. PHOTO: NVCC

Talking about the current situation of online fraud, Mr. Ngo Minh Hieu, Director of Chongluadao.vn project, commented: "Currently, the situation of online fraud in Vietnam is at an alarming level. Compared to countries in the region such as Thailand, Malaysia or the Philippines, Vietnam has a faster rate of cybercrime explosion, especially after the pandemic and the popularity of social networks. Tricks are increasingly sophisticated, from impersonating relatives, government agencies, recruitment fraud, financial investment... to new forms such as Deepfake, AI voice cloning, and impersonation through OTT applications such as Zalo, Telegram. Unlike developed countries, Vietnam still lacks an early warning system, and people are not fully equipped with knowledge about digital security, making it easy for criminals to take advantage of cognitive gaps to attack."

According to expert Ngo Minh Hieu, people need to clearly understand the risks when using technology, especially when sharing personal and financial information. On the part of state management agencies, banks, telecommunications enterprises and the technology community, there must also be cooperation and rapid application of AI technology to manage risks. For example, the banking system should apply AI to detect unusual transactions, the telecommunications industry needs to apply AI to filter fake calls/text messages, and platforms should have transparent fraud reporting channels.

"Regarding the Chongluadao.vn project, we are making efforts to speed up the construction of a centralized database of fraudulent addresses (URLs, account numbers, phone numbers, virtual wallets, etc.) and integrate it into popular platforms. With a real-time user warning system using AI researched by young students, we will integrate it into the Chongluadao.vn website and will officially launch it for free use by the community before the April 30 holiday. This is also a project that we expect will minimize risks for the internet user community and plan to bring it to the "Vietnamese Talent" program this year," Mr. Ngo Minh Hieu shared.

Source: https://thanhnien.vn/4-thanh-nien-tao-cong-cu-chong-lua-dao-truc-tuyen-bang-ai-185250421154003719.htm