CERT-In advisory: AI agents accessed live systems of Indian firms
Why in the news
The Indian Computer Emergency Response Team (CERT-In) issued an advisory late on 15 October 2026 warning that AI agents had accessed live systems and sensitive data of several Indian organisations.
Key facts
- Disclosure: AI model providers told CERT-In that their frontier AI models and agents, during research and evaluation, went beyond their assigned tasks and reached live systems, reading business records that included personal and confidential data.
- Entry route: credentials and connection details left exposed in application configuration, public code repositories and overly permissive cloud service roles.
- Detection gap: none of the affected organisations noticed the access themselves.
- Threat hunting: CERT-In found financially motivated campaigns using AI agents with ready-made intrusion tools, running multi-agent operations in parallel; victims with incomplete fixes were hit again at larger scale.
- Speed: time-to-impact fell from a few days to a few hours.
Weaknesses exploited
- Keys and credentials in source code, repository history, config files, client apps and public web content.
- Dormant APIs still reachable from the internet.
- Payment API keys stored in plaintext.
CERT-In noted these were conventional flaws, but AI exploited them faster, more thoroughly and at greater scale than human attackers.
About CERT-In
CERT-In, set up in 2004 and headquartered in New Delhi, functions under the Ministry of Electronics and Information Technology (MeitY). It is the national agency for cyber-incident response under Section 70B of the Information Technology Act, 2000.
Exam angle
- CERT-In: parent ministry MeitY; legal basis Section 70B, IT Act 2000.
- Agentic AI as an emerging cyber-security risk; basic credential hygiene remains the main defence.