ecosystem announced the launch Project Perception — a next-generation cyber defense platform that uses artificial intelligence to automate threat detection and response processes. The solution was created in response to the growth of cyberattacks using generative AI, which is increasingly being used by attackers to find vulnerabilities, automate attacks, and scale malicious campaigns.
In the company note, that traditional security tools, designed to require human intervention in most processes, are increasingly struggling to cope with modern threats. That's why Project Perception is built around autonomous AI agents that can continuously analyze the environment, identify potential risks, and help eliminate them.
Three types of AI agents
The platform architecture involves the use of three specialized categories of agents:
- Red Team — simulate possible attack scenarios and identify potential vulnerabilities before cybercriminals exploit them.
- Blue Team — analyze the security context, assess the level of risk, and prioritize response.
- Green Team — implement recommended measures to protect and eliminate identified problems.
The collaboration of these agents allows you to automate the full risk management cycle — from identifying a potential threat to implementing protective measures.
Multi-model approach
One of the key features of Project Perception is its multi-model approach. Instead of using one large language model, the platform selects the most efficient AI engine depending on the specific scenario. According to Microsoft, this allows for an optimal balance between accuracy, speed, performance, and processing cost.
As an example, the company cites the MDASH vulnerability management system, which integrates a specialized model MAI-Cyber-1-FlashAccording to Microsoft's internal tests, the new configuration demonstrated 96% effectiveness in the CyberGym test suite, which is 12 percentage points exceeds the Mythos model. In addition, operating costs were reduced by almost 50 %.
Integration with the Microsoft ecosystem
Project Perception integrates with existing Microsoft Security solutions, enabling AI agents to quickly receive incident intelligence, analyze security events, and help neutralize threats. This is expected to reduce response times to attacks and allow cybersecurity professionals to focus more on complex incidents and strategic tasks.
Microsoft also emphasizes that the platform is designed in accordance with the principles of Responsible AI and contains control, management, and audit mechanisms necessary for corporate use.
Public advance access to Project Perception scheduled for August 3, 2026 The new platform is part of Microsoft's long-term strategy to introduce artificial intelligence into cyber defense systems and automate the work of security monitoring centers (SOCs).
