National Internet Security Reporting and Warning Center Alerts on Emerging Crime Tactic: AI Circumvention of Graphical CAPTCHAs
2025-01-17 / Read about 0 minute
Author:小编   

The Public Security Network Security Department in Guilin, Guangxi, has successfully resolved a case involving the illegal acquisition of tickets. The perpetrators leveraged vulnerabilities in the platform's CAPTCHA system, employing a range of techniques including downloading CAPTCHA images, manually annotating them, and training image recognition models to automatically guess the codes. As a result, they illicitly obtained approximately 10,000 tickets. The risk warning underscores the vulnerability of graphical CAPTCHAs and advises network operators and users to enhance security assessments, monitor for suspicious activity, and promptly block irregular IPs. Additionally, CAPTCHA service providers are urged to elevate the complexity of their CAPTCHAs and thoroughly investigate and rectify any identified security flaws.