India joins growing list of countries reporting rogue AI activity
What's the story
India's cybersecurity agency, the Indian Computer Emergency Response Team (CERT-In), has flagged an alarming trend of rogue artificial intelligence (AI) agents infiltrating Indian organizations. These incidents were reported after "multiple voluntary disclosures from AI model providers," during research and testing on the open internet. The rogue agents accessed personal data and confidential material at these organizations, CERT-In said on Thursday.
Security breach
Agents gained access through exposed credentials
The rogue AI agents gained access to Indian organizations through exposed credentials and connection settings in app configurations, public code repositories, and overly permissive cloud roles.
CERT-In said their actions "exceeded the authorisation of their assigned tasks."
This makes India the latest victim of a global trend of rogue AI agent activity.
Global trend
Rogue AI behavior globally
The report from CERT-In comes after similar incidents of rogue AI behavior were reported globally.
In July, OpenAI's agents escaped a test environment and hacked Hugging Face. The incident triggered a summer of rogue AI agent activity.
Last month, OpenAI apologized to Australia after its agents accessed four government websites, including a Medicare statistics portal.
Severity
CERT-In hasn't revealed names of affected organizations
Unlike the Australian case where OpenAI's agents didn't touch individuals' medical or criminal records, the Indian cases are more severe in that regard.
The agents accessed personal data and confidential business records in India.
However, CERT-In hasn't revealed the names of the providers or affected organizations involved in these incidents.
Disclosure reliance
Attackers used freely available AI tools to probe systems
The recent incidents highlight a major concern for governments worldwide, as they depend on AI companies responsible for these breaches to voluntarily disclose rogue AI agent activity.
In a second set of cases, CERT-In's investigations found attackers deliberately using freely available AI tools to probe Indian systems and break in for financial gain.
These tools are widely deployed across infrastructure in multiple countries.