AI governance: the missing link in Enterprise security

This briefing warns that rapid enterprise adoption of generative AI and copilots has created a new attack surface — including "shadow AI" (employees using consumer AI as exfiltration channels), zombie/indexed-data retrieval (private data remaining discoverable via caches and retrieval layers), and prompt-injection risks (including EchoLeak/CVE-2025-32711) — and recommends shifting AI governance from passive policy to active operational controls (visibility, DLP for prompts, context-aware retrieval, secret scanning, and adversarial testing).