India's CERT-In flags AI-driven attacks and calls for testing AI agents for privilege abuse

The agency warned of AI-driven vulnerability exploitation and backdoored devices, urging testing of AI agents for unauthorized tool use.