Microsoft Launches Project Ire, an AI Agent Designed to Autonomously Detect and Classify Malware

Microsoft plans to integrate the Project Ire prototype into Microsoft Defender as a Binary Analyser to enhance threat detection capabilities.

Microsoft unveiled a new artificial intelligence (AI) agent called Project Ire that can autonomously analyze and classify malware. Currently available as a prototype, Project Ire has been tested both in controlled environments and real-world scenarios, demonstrating its ability to fully reverse engineer software without human assistance. The AI performs multi-level analysis to determine whether software is benign or malicious, achieving a high level of precision in an area where AI typically requires human oversight.

According to Microsoft, Project Ire was developed through collaboration among Microsoft Research, Defender Research, and Microsoft Discovery & Quantum teams. The system leverages advanced language models alongside specialized tools designed for binary software analysis, enabling it to perform complex reverse engineering tasks independently.

The company plans to integrate Project Ire into its Microsoft Defender platform as a Binary Analyser to boost threat detection. Defender currently protects over one billion active devices monthly—a scale that poses significant challenges for human analysts. Until now, Microsoft had refrained from relying on AI for such tasks due to the complexity involved in malware detection through reverse engineering.

Project Ire represents a major step forward in automated cybersecurity, offering scalable and precise malware analysis that could significantly enhance Microsoft’s ability to protect users against emerging threats.