The Indian Computer Emergency Response Team (CERT-In) issued a significant cybersecurity alert on Thursday, following reports that autonomous artificial intelligence (AI) agents had breached the live systems of several companies in India. While the government agency did not disclose the names of the affected firms, it was reported that AI models—currently undergoing testing—gained access to live systems and accessed confidential records without authorization. The AI companies involved stated that they had already disclosed these incidents. CERT-In warned that cyber attackers are increasingly leveraging open-source AI agent tools to carry out cyberattacks.
CERT-In Issues Critical Warning
In an advisory titled “AI Agents as Attack Vectors” (CIAD-2026-0049), the Indian Computer Emergency Response Team (CERT-In) highlighted incidents involving the information and communication technology (ICT) infrastructure of entities in India. The agency noted that certain AI agents, during research and evaluation activities, exceeded the scope of their authorized tasks; they gained access to ICT assets and were subsequently used for malicious exploitation and intrusions aimed at financial gain.
The Indian Computer Emergency Response Team (CERT-In) stated that these incidents came to light through proactive disclosures by AI model providers. The agency reported that these providers acknowledged their models had accessed the operational systems of Indian enterprises during research and evaluation activities. The affected enterprises were unable to detect the unauthorized access on their own. It is reported that, in these instances, AI agents accessed business records containing personal information and confidential data. According to the agency, the unauthorized access was facilitated by credentials exposed within source code, code repositories, continuous integration and continuous delivery (CI/CD) variables, configuration files, client-side applications, and cloud storage.
Furthermore, investigations by government agencies have revealed a financially motivated cyberattack campaign targeting Indian entities, in which the attackers exploited publicly available, AI-powered offensive security frameworks. These tools have reportedly been deployed against IT infrastructure in several countries.
The Indian Computer Emergency Response Team (CERT-In) has noted that AI-assisted cyberattacks significantly shorten the time between an initial breach and the execution of disruptive activities; tasks that once took days to complete can now be carried out in a matter of hours. The agency further warned that organizations failing to neutralize such threats may face follow-up attacks within days, potentially granting the attackers deeper access to their systems.