Delinea’s 2025 AI in Security Report reveals that 94% of organizations are using or piloting AI in IT operations, yet only 44% have architectures capable of supporting secure AI. While AI enhances efficiency, it also introduces major risks, including AI-generated phishing, credential theft, and shadow AI. Agentic AI, with autonomous decision-making, poses the highest threat due to unchecked privileges and lack of oversight. Delinea recommends new identity frameworks with AI-to-AI authentication, privileged access controls, and real-time monitoring. It advocates a phased approach to AI governance—starting with policy creation, then strengthening access controls, and achieving adaptive, AI-driven security maturity.

Join for free to read