How Microsoft is strengthening security for the age of AI -May 2026 UK

Microsoft's latest security updates reflect a clear shift: as organisations accelerate AI adoption, security must evolve to protect not just users and devices but agents, data, and identities across a growing ecosystem. 

Broader visibility across your AI estate

Microsoft Purview now extends visibility to third-party AI tools like Claude. This means security and compliance teams can monitor usage, investigate interactions, and better understand how sensitive data is being accessed across AI platforms - all from a central view. 

Data security posture management goes GA

A new Data Security Posture Management (DSPM)  experience in Microsoft Puview is now generally available. It brings discovery, risk assessment, and remediation together into a single workflow - helping teams identify sensitive data and take action faster and at scale. 

Deeper, more flexible investigations

Microsoft has enhanced data security investigations with optical character recognition (OCR)  and custom analysis capabilities. This allows teams to analyse text within images and define tailored investigation rules, expanding visibility into previously hard to detect risks. 

Strengthen identity recovery

Microsoft Entra ID introduces improved account recovery, enabling secure access restoration when users lose authentication methods. This focuses on stronger identity verification and reduces the risk of compromised recovery processes.

Secure environments for AI agents 

Microsoft is also introducing a new approach to managing AI agents securely. Windows 365 for Agents, now expanding in public preview, works alongside Microsoft Agent 365 to provide a controlled execution environment. 

• Agent 365 governs what agents are allowed to do, using organisational policies and identity controls
  

• Windows 365 for Agents defines where agents operate, providing dedicated Cloud PCs for them to run tasks, applications, and workflows
• Together, they enable a fully managed, auditable environment  for AI agents to operate securely

This marks an important step towards treating AI agents as part of the enterprise workforce - requiring the same level of governance, visibility, and control.

What this means for organisations

Security is becoming more integrated, intelligent, and AI-aware. Microsoft's latest updates focus on:

• Extending visibility across AI and third-party tools
• Simplifying how security teams identify and respond to risk
• Strengthening control over sensitive data and identities
• Supporting governance as AI becomes embedded in everyday work. 

For organisations adopting Copilot and other AI tools, these changes are a step towards managing AI securely without slowing innovation. 

Need support securing your environment?

As AI and cloud ecosystems grow, so do the security challenges. If you'd like guidance on how these updates apply to your organisation, get in touch with the Codec team for a tailored security consultation

Reach out to: info@codec.ie