AI Agents Are the New Insider Risk: A GRC Playbook for New York’s Financial Sector
AI agents are already inside major financial institutions, quietly approving transactions, touching sensitive data, and interacting with customers at machine speed. Most firms have no clear inventory of how many agents are running, what permissions they inherited, or whether anyone is actually watching them. NYDFS has already said Part 500 applies right now. This article reveals the identity gaps, real incidents, vendor‑embedded agents, and governance failures already unfolding, and the playbook institutions need before an examiner asks who was watching the agent.











