Microsoft expands its AI security and governance offerings to secure the 'agentic workforce' where AI agents and humans collaborate.
New capabilities like Entra, Defender, and Purview are integrated into Azure AI Foundry and Copilot Studio to secure AI apps across the development lifecycle.
Entra Agent ID is launched to manage identities of AI agents, while Defender for Cloud security insights are now integrated into Azure AI Foundry for AI-specific threat alerts.
Purview platform gets a new SDK for policy enforcement and data loss prevention in AI systems, and Azure AI Foundry introduces security updates like 'Spotlighting' to detect prompt injection attacks.