TLDR: Microsoft has introduced a new, comprehensive AI Agent Framework, including services like Azure AI Foundry Agent Service and Microsoft Entra Agent ID, designed to unify, govern, and scale AI agents across enterprise environments. This strategic shift, highlighted at Build 2025, aims to transform AI agents into foundational components of enterprise infrastructure, enabling secure, collaborative, and autonomous operations while integrating existing tools like AutoGen.
Microsoft is spearheading a significant evolution in enterprise artificial intelligence with the introduction of a robust AI Agent Framework, positioning AI agents as a core layer of future enterprise infrastructure. Unveiled at its Build 2025 conference, this strategic initiative aims to unify, govern, and scale AI agents, addressing critical needs for security, manageability, and operational efficiency within complex business environments.
The centerpiece of this new framework is the Azure AI Foundry Agent Service, now generally available, which provides a platform for developers to build, manage, and scale AI agents that automate business processes. This service is designed to support multi-agent workflows, allowing specialized agents to coordinate on high-level tasks. Notably, the Foundry Agent Service features a unified runtime that merges the Semantic Kernel SDK and the AutoGen framework, indicating an integration and evolution of AutoGen rather than its retirement, as it now forms a component of Microsoft’s broader agent strategy.
To ensure the secure deployment and management of these autonomous entities, Microsoft introduced Microsoft Entra Agent ID. This new identity and governance layer extends existing IT visibility, access control, and lifecycle policies to non-human software agents. Through the Entra admin portal, enterprises can apply Conditional Access, auditing, and lifecycle rules to agent identities, ensuring robust governance across agents built with Microsoft services like Copilot Studio, Azure AI Foundry, and Security Copilot. Microsoft also plans to support third-party agent integration, collaborating with companies like ServiceNow and Workday to extend governance into multi-vendor workflows.
The company’s vision extends to fostering an ‘open agentic web,’ where AI agents operate autonomously and collaborate across platforms with minimal human intervention. This is supported by initiatives such as NLWeb, an open protocol enabling websites and applications to easily implement conversational AI features, and support for open standards like Agent2Agent and Model Context Protocol (MCP) to ensure interoperability across diverse agent frameworks.
Further enhancing the ecosystem, Microsoft announced plans for an Agent Factory, a framework to streamline the creation and deployment of AI agents across platforms like Microsoft 365 and Azure. The AgentOps tools provide advanced monitoring and optimization capabilities, while Copilot Tuning offers low-code customization for Copilot agents. For scientific research, the new Microsoft Discovery platform leverages AI agents to automate steps from hypothesis generation to data analysis.
Also Read:
- Microsoft Fortifies Sentinel with Advanced AI Agent Capabilities to Combat Evolving Cyber Threats
- Leading AI Agents of 2025 Revolutionize Business Productivity and Workflows
Industry analysts suggest this comprehensive approach could significantly reduce friction in AI solution implementation, shorten development cycles, and enhance operational efficiency by enabling interconnected AI systems with persistent memory. Microsoft’s announcements at Build 2025 underscore a fundamental shift towards treating AI agents as composable, governed, and integrated components of the enterprise stack, poised to revolutionize how businesses utilize artificial intelligence.
