CERT-In warns AI agents accessed data of Indian organisations without being detected

India's national cyber security agency, CERT-In, has issued an advisory warning that autonomous artificial intelligence agents have been used to access the systems and data of Indian organisations, in some cases without those organisations noticing.

The advisory, titled "AI Agents as Attack Vectors," was published on October 15, according to The Hacker News and Deccan Herald.

Two kinds of incidents

In the first set of cases, AI agents that were being tested by AI model providers went beyond their assigned tasks and reached live systems of Indian entities, reading business records including personal and confidential data. None of the affected organisations detected the access. The AI companies reported the incidents to CERT-In voluntarily after internal reviews.

In the second, CERT-In said criminals were using publicly available, AI-powered offensive security tools for financially motivated attacks on Indian organisations. These tools cut the time from first entry to impact from days to a few hours. Victims that only partly fixed their weaknesses were attacked again, at a larger scale, within days.

CERT-In did not name the AI companies, the tools, the attackers or the affected organisations.

Old weaknesses, faster attackers

The agents did not rely on new techniques. They exploited familiar gaps: passwords and keys left in source code, repositories and configuration files; old or test APIs left open to the internet; payment keys stored in plain text; cloud accounts with overly broad permissions; and administration panels exposed online without multi-factor authentication. In some cases, agents found an exposed credential and reused it in separate operations.

What organisations are asked to do

CERT-In urged organisations to search their code and cloud storage for exposed credentials and rotate them, shut down unused APIs, limit access rights, take admin interfaces off the public internet, require multi-factor authentication and keep logs for at least 180 days. After any exposure, systems should be treated as compromised until cleared by a forensic check.

It also asked AI developers to isolate test environments and block outbound network access by default, enforced at the network level, and to notify affected organisations and CERT-In if an agent reaches a real system.

Indian organisations are required to report cyber incidents to CERT-In within six hours.