The new system is designed to operate as both an attacker and defender, using AI to identify vulnerabilities before they can be exploited.
Microsoft has launched an AI cybersecurity system designed to detect software weaknesses before attackers can use them. The system acts in dual roles, functioning as both an attacker and a defender to surface vulnerabilities earlier in the security process. The approach reflects a broader push in cybersecurity to use AI offensively for simulated attacks and defensively for protection, helping organizations test systems before real-world hackers exploit gaps.