News from the AI & ML world
DeeperML
Alex Simons@Microsoft Security Blog
//
Microsoft is heavily invested in the future of AI agents, showcasing significant advancements at Microsoft Build 2025. The company envisions an "age of AI agents" where systems proactively assist users, spotting problems and suggesting solutions independently. This shift moves beyond simple request-response interactions to a more dynamic and helpful AI experience. To support this evolution, Microsoft is focusing on adapting and enhancing existing identity standards, particularly OAuth, to manage how these agents access data and operate across diverse connected systems, spanning from APIs to sensitive business processes.
Microsoft is introducing Microsoft Entra Agent ID, extending its identity management and access capabilities to AI agents. This is a crucial step towards securing the "agentic workforce." Furthermore, the company has already launched a public preview of the Conditional Access Optimizer Agent, a multi-functional AI agent capable of analyzing Conditional Access policies, identifying security gaps, and deploying new policies in pilot mode. Similar investments are being made in agents for developer and operations workflows, such as SWE and SRE agents, designed to boost productivity for teams involved in application development and maintenance.
Microsoft emphasizes that the industry needs to continue collaborating on establishing identity standards for agent access across various systems. The developments reflect Microsoft's commitment to building a robust and sophisticated suite of agents designed to augment and amplify organizational capabilities. Ultimately, Microsoft envisions these AI agents as integral components of an open agentic web, which requires a strong foundation of Zero Trust security principles.
References :
Microsoft is introducing Microsoft Entra Agent ID, extending its identity management and access capabilities to AI agents. This is a crucial step towards securing the "agentic workforce." Furthermore, the company has already launched a public preview of the Conditional Access Optimizer Agent, a multi-functional AI agent capable of analyzing Conditional Access policies, identifying security gaps, and deploying new policies in pilot mode. Similar investments are being made in agents for developer and operations workflows, such as SWE and SRE agents, designed to boost productivity for teams involved in application development and maintenance.
Microsoft emphasizes that the industry needs to continue collaborating on establishing identity standards for agent access across various systems. The developments reflect Microsoft's commitment to building a robust and sophisticated suite of agents designed to augment and amplify organizational capabilities. Ultimately, Microsoft envisions these AI agents as integral components of an open agentic web, which requires a strong foundation of Zero Trust security principles.
Share:
References :
- blogs.microsoft.com: At Microsoft Build 2025, we’re showing the steps we’re taking to make the AI agent era a reality through our platforms, products, and infrastructure. The post appeared first on .
- Microsoft Security Blog: Our industry needs to continue working together on identity standards for agent access across systems. Read about how Microsoft is building a robust and sophisticated set of agents. The post appeared first on .
- www.microsoft.com: At Microsoft Build 2025, we’re taking important steps to secure the agentic workforce. We are excited to introduce Microsoft Entra Agent ID which extends industry-leading identity management and access capabilities to AI agents. The post appeared first on .
Classification:
- HashTags: #Microsoft #AI #ZeroTrust
- Company: Microsoft
- Target: Workforce
- Product: AI
- Feature: AI Agent
- Type: AI
- Severity: Informative