India's CERT-In issued an advisory on October 15 saying AI agents reached live systems through exposed credentials and over-permissive cloud roles, according to the source report.
The advisory points to access controls rather than the models themselves as the weak point.
Source: techcrunch.com, reported 10-16. https://techcrunch.com/2026/10/16/india-becomes-the-latest-country-to-report-rogue-ai-agent-activity/