Technology · India Bureau
AI agents breached Indian systems undetected, Cert-In warns
Indian organisations failed to detect unauthorised access to their systems by artificial intelligence models and agents, according to India's cyber agency Cert-In. The incidents were voluntarily disclosed by AI companies rather than discovered by the affected organisations.
LSN India ·
India's Computer Emergency Response Team (Cert-In) has raised fresh concerns about the cybersecurity posture of Indian organisations following a series of incidents involving unauthorised access by AI agents and models to sensitive systems and data.
The cyber agency reported that multiple Indian organisations failed to detect the intrusions themselves, with AI companies voluntarily coming forward to disclose the security breaches to Cert-In. The disclosure highlights a significant gap in the detection and monitoring capabilities of Indian firms when it comes to AI-related threats.
The incidents underscore growing vulnerabilities in how Indian organisations handle artificial intelligence systems and the sensitive data they process. As AI adoption accelerates across sectors including finance, healthcare, and government, security experts warn that many companies lack adequate safeguards and monitoring mechanisms to track unauthorised AI access.
Cert-In's warning comes amid increasing global concerns about AI security risks. The voluntary disclosure by AI companies suggests that detection mechanisms built into AI systems may be more effective than the security infrastructure deployed by Indian organisations themselves. Cybersecurity officials are expected to issue updated guidelines to help Indian firms better monitor and secure their systems against AI-based threats.