CERT-in Issues Critical Warning After AI Agents Broke Into Indian Organisations' Live Systems

The Indian Computer Emergency Response Team (CERT-In) issued a critical cybersecurity advisory on Thursday after autonomous artificial intelligence (AI) agents reportedly accessed live systems of Indian companies. The government agency did not name the affected organisations, but the AI models under test claimed to have broken into thier live systems and read confidential records without authorisation. The AI companies involved said they had disclosed the incidents. CERT-In warned that threat actors are increasingly using open-source agentic AI tools to carry out cyberattacks.

CERT-In Issues Critical Warning

In its advisory CIAD-2026-0049, titled ‘AI Agents as Attack Vectors', CERT-In highlighted incidents involving the information and communication technology (ICT) infrastructure of Indian entities. The agency noted that some AI agents crossed the boundaries of their authorised tasks during research and evaluation activities, gaining access to ICT assets, exploitation and financially motivated intrusions.

CERT-In claims that the incidents came to light through voluntary disclosures by AI model providers. It states that AI model providers admitted that their models, during research and evaluation activities, had reached the live systems of Indian organisations.  The affected organisations could not detect the unauthorised access on their own. During the incidents, the AI agents reportedly accessed business records containing personal information and confidential data. The agency says the unauthorised access was facilitated by exposed credentials in source code, repositories, continuous integration and continuous delivery (CI/CD) variables, configuration files, client applications and cloud storage.

Additionally, investigations by the government agency found a financially motivated campaign against Indian entities in which cyber threat actors abused publicly available AI-native offensive-security frameworks.  These tools were reportedly being used across IT infrastructure in several countries.

CERT-In states that the AI-assisted cyberattacks have significantly reduced the time between an initial breach and the execution of damaging activities, with some attacks that previously took days now being completed within hours. The agency also warned that organisations that fail to remove the threat can face repeat attacks within days, with attackers gaining deeper access to their systems.

CERT-In's warning highlights the importance of restricting the access granted to AI agents, securing credentials, and reviewing cloud permissions to reduce the exposure of confidential data.