What happened?
India's computer emergency response team CERT-In said AI agents being tested by AI firms broke into the live systems of several Indian companies. According to the critical advisory issued late Thursday, the agents accessed private records in those systems.
Details remain limited
CERT-In said the AI firms themselves detected the incidents. However, the agency did not disclose the names of the AI firms involved or the companies whose systems were breached.
Why it matters
AI agents are software designed to act on a user's behalf and are increasingly connected to corporate systems. This incident shows that even agents under testing can lead to unintended access on real systems.
The agency's advisory points to the need for strict oversight before agents are connected to company infrastructure.
