Microsoft announced a suite of security updates and enhancements aimed at securing agentic AI. The company introduced Microsoft Agent 365, a control plane for centralized visibility and governance over AI agents. Available on May 1, it includes Microsoft Defender, Entra, and Purview capabilities to manage agent access, reduce data oversharing, and defend against emerging threats.
The updates also include enhanced identity protection, data security, and cloud and endpoint defense. New features like Entra Backup and Recovery, Entra Tenant Governance, and Entra external multifactor authentication are now in preview or generally available.
Microsoft is also expanding its security stack with new capabilities such as Entra Internet Access prompt injection protection, enhanced container security, and predictive shielding. These features aim to provide comprehensive and adaptive security for AI-driven environments.
Microsoft Sentinel, a key component in the company's AI-driven security strategy, is being upgraded to serve as an agentic defense platform, unifying context, automating workflows, and standardizing governance across security tools.
Microsoft will showcase these new features at the RSAC Conference in San Francisco from March 23-26.