Microsoft has launched its first cybersecurity-focused artificial intelligence model, MAI-Cyber-1-Flash, alongside a new security platform called Perception. This move positions Microsoft as an aggressive competitor to established players like Anthropic and Google.

The MAI-Cyber-1-Flash is integrated with MDASH, Microsoft’s software vulnerability identification tool. According to Mustafa Suleyman, CEO of Microsoft AI, it outperforms other benchmarked models on Cyber Gym and will be rolled out immediately into production.

Perception, the new platform, deploys teams of agents to automate various security workflows, from identifying bugs to remediation. Blue teams detect existing vulnerabilities, red teams simulate potential attacks, while green teams take corrective actions. Dave Weston, lead engineer for Perception, describes it as a significant efficiency upgrade for corporate defenders.

Microsoft’s venture into AI cybersecurity comes amid increased use of AI by hackers. ‘We’re shipping this into production immediately,’ Suleyman said, highlighting the importance of defending against AI with AI at scale and speed.

This launch marks another step in the tech giants’ race to secure digital assets in a world where cyber threats are becoming increasingly sophisticated and automated.