Shadow AI: Managing the Security Risks of Unsanctioned AI Tools | UpGuard

**Executive summary:** The report analyzes the growing problem of 'Shadow AI'—employees using unapproved generative AI tools—and describes risks including uncontrolled data exfiltration, regulatory and IP exposure, malicious or insecure AI services, and gaps in traditional IT asset management; it recommends user-centric monitoring (network traffic, browser/extension, SaaS logs, endpoint analysis), cataloging and risk-rating tools, adaptive policies, and ongoing employee education to enable safer AI adoption.