Cybersecurity firms are leveraging big data to construct anti-fraud detection models, while overseas fraud syndicates are also upgrading their fraudulent tactics using AI technology. This makes the prevention of cross-border fraud more challenging.
Recently, Aviram Ganor, General Manager of Riskified for Europe, the Middle East, Africa (EMEA), and Asia-Pacific, spoke to a reporter from China Business News, stating that data indicates fraudsters are employing AI technology to enhance their fraudulent methods, such as creating infinitely iterated fake accounts, making identification and prevention more difficult.
Information security risks in overseas ticket purchases
Statistics from the National Immigration Administration show that between July and August, national border inspection authorities facilitated 110 million entries and exits of Chinese and foreign personnel, averaging 1.779 million per day, a 30% increase compared to the same period last year, and a 13% increase compared to May and June. On August 24th, the historical second-highest level of 2.237 million entries and exits was reached.
In 2024, outbound tourism has significantly rebounded, leading to a phenomenon of cross-border fraud in air ticket payments. Data shows that the risk of aviation ticket orders in the first half of the year increased by 14% year-on-year. During the peak travel season, the order fraud rate reaches its peak, with the fraud rate in August being 40% higher than the average of other months.
Due to the concentrated travel of Chinese passengers in a short period and the common practice of booking tickets overseas, language and regulatory differences may lead to transactions being mistakenly rejected, resulting in the theft of personal information.
Aviram Ganor introduced several types of fraud to the reporter. "Passenger information may be exploited by overseas fraudsters, using stolen or cloned credit cards to book travel—since fraudsters do not spend their own money when purchasing tickets, they can sell tickets at a discount—either by selling them to illegal customers through dark web forums or by selling them to unsuspecting customers through seemingly legitimate fake travel websites, and profiting from all the proceeds."
Fraudsters also engage in chargebacks or refusals to pay for high-priced tickets, often requesting refunds to another payment method to convert the ticket funds into cash. In addition, criminals may take over accounts—when they can access the accounts of legitimate customers on travel websites, they can redeem rewards or miles and sell them for profit.
The current solution is to use AI to establish fraud detection models. Aviram Ganor said, "Algorithms will continuously adjust and optimize based on fraud trends. By observing behavioral patterns in merchant data, algorithms can detect anomalies and suspicious activities without preset strict rules. This means that when risk patterns emerge, merchants can discover and protect themselves in real-time."
Furthermore, he added, "Algorithms can use large datasets from multiple retailers and millions of transactions to identify risk trends and prevent fraudsters from reaching more unsuspecting retailers before they do."Fraudsters can analyze large amounts of data
As corporate network security technology upgrades, AI tools enable fraudsters to expand their operations, making it easier to commit fraud on a global scale. Aviram Ganor said: "AI helps fraudsters analyze massive amounts of data, create fake accounts, or carry out covert fraudulent transactions, often causing losses before users realize it."
This poses challenges to network security companies and requires businesses to continuously improve their network defense capabilities and fraud prevention measures. Aviram Ganor believes: "AI technology will further drive the evolution of cross-border payment fraud patterns, bringing more intelligent, complex, and rapid attacks."
It is understood that fraudsters may use AI technology to amplify certain methods, such as creating an infinite number of fake accounts through iteration, making identification and prevention more difficult. Fraudsters may also adopt more complex attack patterns, generating a large amount of simulated fake transaction data to confuse anti-fraud systems. The application of AI also allows them to adjust their attack strategies more quickly, exacerbating the threats faced by cross-border payment systems.