Microsoft has introduced a new artificial intelligence model designed specifically for Cybersecurity. The company also unveiled a new AI-powered platform that uses teams of intelligent agents to automate security operations and help organizations respond to cyber threats more quickly.
The new AI model is called MAI-Cyber-1-Flash. It has been developed to detect difficult software vulnerabilities in large and complex codebases before attackers can exploit them. Microsoft says the model is now being used through its existing MDASH security system.
According to the company, MDASH is an AI-driven platform that identifies, verifies, and helps fix software vulnerabilities. It combines multiple AI models with more than 100 specialized AI agents. Each agent performs a specific task to improve the overall Cybersecurity process.
Microsoft explained that MAI-Cyber-1-Flash can complete up to 90 percent of the tasks handled by MDASH. More advanced AI models, including GPT-5.4, are then used only for the most difficult security challenges. This approach helps reduce computing costs while maintaining strong security performance.
The company said the new system is designed to continuously scan software for weaknesses. Finding vulnerabilities early allows developers to fix problems before hackers have a chance to exploit them.
Microsoft also announced Project Perception, a broader AI security platform that goes beyond finding software bugs. The platform uses teams of AI agents to monitor systems, investigate threats, and respond automatically when security risks are detected.
Project Perception organizes its AI agents into three groups known as red, blue, and green teams. Each team has a different role in protecting computer systems.
Red-team agents act like attackers. They search for possible weaknesses that hackers could use to gain access to systems. Their goal is to identify vulnerabilities before real cybercriminals discover them.
Blue-team agents examine suspicious activity and review security alerts. They analyze available information to determine whether a threat is genuine or a false alarm. This helps security teams focus on the most serious risks.
Green-team agents are responsible for taking corrective action. They help strengthen security protections and recommend or create fixes for identified vulnerabilities. Together, the three teams create a continuous cycle of detecting, evaluating, and resolving security issues.
Microsoft said the platform can significantly reduce the time required to respond to threats. According to the company, tasks that previously took several Cybersecurity experts and many hours of manual work can now be completed within minutes.
The company also shared performance results for its new technology. Microsoft stated that MDASH, powered by MAI-Cyber-1-Flash and GPT-5.4, achieved a score of nearly 96 percent on CyberGym. CyberGym is a benchmark used to measure how effectively AI systems detect real software vulnerabilities.
Microsoft claims the new system performed better than several competing AI security solutions. The company also said the updated setup reduces operating costs by around 50 percent compared to its previous MDASH configuration.
Microsoft Security officials explained that these new tools are becoming increasingly important because attackers are also using artificial intelligence. AI allows cybercriminals to search for software weaknesses much faster than traditional methods.
The company believes businesses need equally advanced defensive tools to keep pace with AI-powered attacks. Microsoft said its global security systems currently process more than 100 trillion security signals every day across devices, cloud services, networks, identities, and applications.
Microsoft is entering a rapidly growing market for AI-powered security solutions. Other technology companies have also introduced similar platforms that use artificial intelligence to detect vulnerabilities and improve digital protection.
In other news read more about Microsoft Edge Restores Background YouTube Playback in Experimental Update
Project Perception will enter public preview on August 3, while MAI-Cyber-1-Flash is already being deployed through MDASH. Microsoft says these new technologies represent an important step toward faster, more automated, and more effective Cybersecurity for businesses worldwide.
