AI agents being tested by developers accessed live systems belonging to Indian organisations and read confidential business records, according to India’s cybersecurity agency, raising concerns about the security risks associated with advanced AI models.
The Indian Computer Emergency Response Team (CERT-In) said AI providers voluntarily reported the incidents after their systems exceeded their authorised tasks. The agency did not identify the companies involved or the affected organisations.
In separate investigations, CERT-In also found that cybercriminals were using freely available AI tools to breach computer systems for financial gain. The agency said the technology had accelerated some attacks, reducing the time from initial access to damage from days to just a few hours.
The findings highlight the need for stronger security controls as AI capabilities expand and become increasingly accessible to both businesses and malicious actors.
Leave a comment