Microsoft Perception Is Now Available, Using AI Agents to Detect and Fix Security Threats
Microsoft has announced the general availability of Microsoft Perception, an AI-powered cybersecurity platform that uses autonomous agents to detect, investigate, and address security threats.
The company announced the launch at Microsoft Ignite 2026, expanding a platform previously known as Project Perception, which Microsoft first introduced in July 2026.
According to Microsoft’s announcement, Perception combines several AI security agents to help organizations identify vulnerabilities, investigate suspicious activity, and resolve security issues.
Microsoft Perception uses three types of AI security agents
Microsoft Perception coordinates three types of AI agents, each responsible for a different part of the security process.
- Red team agents simulate cyberattacks to identify vulnerabilities before attackers can exploit them.
- Blue team agents monitor threats, investigate incidents, and analyze security alerts.
- Green team agents address vulnerabilities and help fix security problems.
The agents work together through automated playbooks, allowing organizations to continuously identify, investigate, and respond to security threats.
Microsoft Perception also integrates with existing Microsoft security products, including Defender, Entra, Intune, Sentinel, and Purview.
Microsoft adds automated playbooks and vulnerability shielding
With general availability, Microsoft Perception introduces several capabilities designed to improve threat detection and remediation.
These include actionable security plans that help organizations prioritize threats, new automated playbooks that coordinate AI agents, and vulnerability shielding.
The vulnerability shielding feature, currently in preview, provides temporary protection for vulnerable systems while organizations develop and deploy permanent fixes.
Microsoft also plans to introduce additional integrations with Microsoft Purview over the coming months.
Microsoft’s MDASH AI code scanner now supports larger repositories
Alongside Perception, Microsoft has expanded the preview of Codename MDASH, its AI-powered code scanner designed to identify software vulnerabilities.
The updated scanner can now analyze software repositories up to 5 GB in size, allowing organizations to scan larger codebases for potential security issues.
Microsoft has also added support for regulated cloud environments, making MDASH accessible to organizations with stricter security and compliance requirements.
The announcements are part of Microsoft’s broader effort to strengthen security across its products. At Ignite 2026, the company also introduced new Windows kernel protections and faster security updates.
Other announcements include the Windows Server 2028 public preview, which brings native NVMe support and post-quantum security capabilities, and the launch of seven new Microsoft AI models.
Read our disclosure page to find out how can you help Windows Report sustain the editorial team. Read more
User forum
0 messages