AI Tools Exploit Vulnerabilities in CBSE Portals, Prompting Security Review

A panel from IIT Madras and IIT Kanpur discovered that AI tools, including Claude, were used to identify vulnerabilities in the CBSE's on-screen marking (OSM) portal. This breach raised concerns about the security knowledge of the vendor responsible for the system. Following the findings, sensitive data was transferred to a government-controlled AWS segment to mitigate risks. The review was initiated due to complaints from students regarding issues with answer sheet scans and portal performance. The IIT experts are conducting a comprehensive evaluation of CBSE's IT ecosystem, with a report expected to recommend improvements. The incident has drawn attention from cybersecurity agencies like CERT-In and the Ministry of Electronics and Information Technology. The scope of the impact extends to other examination platforms, including JEE Advanced, amid ongoing security checks.

Key Points: • AI tools, including Claude, exploited vulnerabilities in CBSE's OSM portal. • Sensitive data was moved to a government-controlled AWS segment for security. • A comprehensive review of CBSE's IT systems is underway, involving IIT experts.