Microsoft offers Security Copilot, a generative AI assistant for security and IT professionals, generally available worldwide since April 2024. It processes natural language questions from analysts and returns findings drawn from Microsoft's global threat intelligence to help teams detect, investigate, and respond to security incidents. In March 2025, Microsoft announced six autonomous AI agents for Security Copilot, available for preview in April 2025, designed to handle high-volume tasks such as phishing triage, access policy optimization, and vulnerability remediation without requiring step-by-step human input.

Announced in March 2023, Security Copilot reached general availability on April 1, 2024 with pay-as-you-go pricing of $4 per hour per Security Compute Unit. It integrates with Microsoft's full security product portfolio including Defender XDR, Sentinel, and Microsoft Entra. The AI processes natural language questions and returns findings that would otherwise require hours of manual investigation — for example, tracing how an attacker moved through a network. In March 2025, Microsoft unveiled six autonomous AI security agents capable of independently triaging phishing alerts, flagging identity risks, and optimizing access policies. Microsoft's internal studies found that security analysts were 22% faster and 7% more accurate with Security Copilot than without it. As of November 2025, it is included at no additional cost for Microsoft 365 E5 customers.