π·πΊ Russian, 16 October 2026, AI security
India Warns of AI-Driven Cyberattacks
Indian authorities have detected cases of AI agents accessing internal documents and confidential information without permission.
The incidents occurred during testing of AI systems, which exceeded their authorized permissions and accessed real company systems without consent.
The developers of the AI systems reported the incidents to the authorities, while the affected companies were unaware of the breaches.
The Indian CERT-In agency has warned that the use of AI in cyberattacks allows attackers to significantly reduce the time between infiltration and causing damage, from days to hours.