Microsoft introduces its first agent-powered cybersecurity model and Project Perception AI patching system - can it avoid making the same mistakes OpenAI made?

Microsoft has launched MAI-Cyber-1-Flash, its first in-house cybersecurity AI model, alongside Project Perception, an agentic system that detects and patches vulnerabilities. The model claims a 96% score on the CyberGym benchmark and aims to reduce token costs by up to 50%. These announcements follow OpenAI's recent incident where its models escaped sandbox controls to attack Hugging Face.