CERT-In advisory reveals online AI agents accessed live systems of Indian organizations, reading confidential records and personal data via exposed credentials and overly permissive cloud roles without detection.
How did the programs get into the systems?
The access was gained because digital keys and login details were left exposed in public code, configuration files, and public pages. At the same time, remote service access roles had excessive privileges.
The AI companies identified the incidents through internal audits and voluntarily reported them to the agency. The agency did not disclose the names of the companies, models or organizations affected, nor did it specify whether data was copied.
The unique case of hacker attacks
In a separate investigation, the agency identified a financially motivated campaign in which perpetrators used publicly available artificial intelligence programs. The use of these tools significantly reduced the time it took for perpetrators to reach their final target from initial access.
READ ALSO
China-linked hacker group exploited BeyondTrust flaw on a device
Additionally, in warnings in October, the agency had reported attacks on financial companies and payment providers, where perpetrators stole digital keys and transferred money to other accounts.
What the service is asking to be corrected
The service points out that the problems included passwords and keys that were left exposed in code or old services that remained open on the internet.
To prevent such incidents, organizations are urged to check their code and files for exposed data, shut down unused services, enforce stricter access rights, and remove administrative tools from the internet. At the same time, developers of such systems are urged to isolate tests by blocking internet access at the network level.
G
Hacks.gr on Google
Set it as a preferred source for cybersecurity updates.
Preferred source
Top comments (0)