AI Vietnam detects fraudulent websites in just 30 seconds, 98% accurate

The AI-based phishing website detection tool developed by a group of students from Swinburne Vietnam University, under the guidance of cybersecurity expert Ngo Minh Hieu (Hieu PC, co-founder of the Vietnam Cyber Fraud Prevention Project), has won the third prize at the Vietnam Talent Awards. This application has the ability to deeply analyze the structure and behavior of websites to detect signs of fraud in real time, giving quick results in under 30 seconds with an accuracy of up to 98%.

The AI-based phishing website identification tool developed by a group of young authors from the Anti-Fraud Company won third prize in the Promising Digital Technology product system.

Fast results under 30 seconds with up to 98% accuracy.

According to the group representative, users can easily check the safety level of any website simply by pasting the link into the system at https://ai.chongluadao.vn. The application works completely online, no installation required, just enter the link (URL) suspected of fraud, the artificial intelligence system will conduct real-time analysis and return results in just a few seconds.

A group of students developed a tool to filter fraudulent websites https://ai.chongluadao.vn

The tool uses next-generation AI (large language model – LLM) to comprehensively evaluate many elements of the website such as links, content, interface, source code and interactive behavior, thereby providing a risk score on a scale of 10 with detailed explanations, with an accuracy of more than 98%.

Not only does it stop at scoring, the system also compiles reports from reputable cybersecurity organizations and displays live video recordings of the page interface, helping users to observe visually without having to access it themselves, avoiding the risk of malware infection.

Notably, this platform is being expanded with a PHP malware scanning feature for WordPress sites at https://wpcheck.chongluadao.vn, helping to detect and prevent early malicious code from being inserted into the system.

Compared to traditional tools that rely solely on blacklists or domain characteristics, this AI solution does not simply conclude whether a site is “safe” or “dangerous” but also clearly explains the reasons, provides evidence and multi-source reports to bring the most comprehensive, transparent and understandable perspective to Internet users.

An impressive point is that this application has the ability to deeply analyze the structure and behavior of the website to detect signs of fraud in real time, giving quick results under 30 seconds with up to 98% accuracy.

“Digital Shield” identifies fake websites

Cybercrime in Vietnam is growing at an alarming rate, causing losses of tens of thousands of billions of VND each year. According to the National Cyber Security Report 2024, on average, one in every 220 smartphone users is a victim of online fraud, with estimated losses of up to 18,900 billion VND (equivalent to 740 million USD).

In Ho Chi Minh City alone, in the first 9 months of 2024, the City Police received 461 cases of online fraud, with a total amount of money stolen of about 982 billion VND.

According to experts, cybercrime is becoming increasingly sophisticated, constantly changing its forms of fraud from impersonating banks and government agencies to taking control of social network accounts. More worryingly, children and teenagers have become the most vulnerable group in the digital age.

This is the reason why the Anti-Fraud team led by cybersecurity expert Ngo Minh Hieu (Hieu PC) has developed a tool to identify fraudulent websites using artificial intelligence (AI), capable of deep analysis and detecting signs of fraud in real time.

This is one of the first tools in Vietnam to help identify fraudulent websites in real time completely automatically using artificial intelligence, integrating a fine-tuned large language model and a thinking model.

“The AI tool to identify fraudulent websites was developed with the goal of providing a “digital shield” to help people easily check and identify fake websites early, contributing to protecting the personal information security and digital trust of Vietnamese people,” the group said.

According to expert Ngo Minh Hieu, the group includes many third-year university students who have taken on key technical roles. The system was developed in 4 months (from January to April 2025), operating based on a network of cyber security intelligence, allowing early scanning and detection of suspicious websites before they harm users.

Immediately after its launch, the tool attracted more than 1,000 analyzes per day, and in less than 4 months, the number exceeded 68,000. Each analysis is a user assisted in protecting information and assets online.

The tool has also been extended with a PHP malware scanning feature for WordPress websites, which has been successfully tested in Vietnix's system with more than 100,000 customers. With a simple, easy-to-use interface, the ai.chongluadao.vn platform helps anyone from ordinary users to businesses to check, analyze and detect risks in just a few seconds.

“Positive feedback from users is the biggest motivation for us to continue improving the tool,” expert Ngo Minh Hieu shared, adding that the group’s goal is to build a safer and more transparent cyberspace for all Vietnamese people.

However, cybersecurity experts recommend that users still need to be vigilant when accessing websites, especially those that require personal information, bank accounts or OTP codes. Although artificial intelligence can support analysis with high accuracy, risk awareness and self-protection skills are still the first and most important "shields" against online fraud.

Source: https://khoahocdoisong.vn/ai-viet-phat-hien-website-lua-dao-chi-sau-30-giay-chinh-xac-98-post2149064631.html