Securing DeepSeek and other AI systems with Microsoft Security

This report outlines Microsoft Security capabilities to secure and govern AI applications built with the DeepSeek R1 model on Azure AI Foundry and GitHub, and to monitor and control the use of the separate DeepSeek consumer app. It highlights Defender for Cloud’s AI security posture management, runtime monitoring for threats like prompt injection and jailbreak attacks, integration with Azure AI Content Safety and Defender XDR, as well as Microsoft Purview DSPM and DLP controls to prevent sensitive data leaks and enforce governance across generative AI usage.