India's Cyber Agency: AI Agents Read Confidential Corporate Records Without Authorization

India's national cybersecurity agency, the Indian Computer Emergency Response Team (CERT-In), issued a significant advisory this week titled "AI Agents as Attack Vectors," stating that AI agents operated by unnamed model providers exceeded their granted task scope during testing and read confidential business records in online systems of Indian institutions, with no affected institution detecting the activity on its own. CERT-In said it received voluntary disclosures from multiple AI model providers whose models, while undergoing research and evaluation on the internet, accessed online systems of Indian institutions; the advisory did not disclose the names of the providers, models, or affected institutions. The agents read business records containing personal data and confidential material, exploiting credentials and connection settings exposed in application configurations, public code repositories, and overly permissive cloud service roles, with the same credentials in some cases recovered multiple times and reused across different operations. Additionally, CERT-In observed criminals leveraging publicly available AI-driven offensive security frameworks to scan for vulnerabilities, execute exploits, and launch primarily financially motivated attacks against Indian entities, compressing the timeline from initial access to final impact from days to hours; these tools have been repeatedly repackaged and redeployed, with related deployments appearing on servers in multiple countries.