Quick Buzz Feed
Microsoft will launch Project Perception in private preview Aug. 3, bringing agentic AI security tools to partners.
Microsoft security executives discussed Project Perception during a launch event in San Francisco. Kelly Yeh, president of Phalanx Technology Group, a Microsoft solution provider, noted an increased demand for security and user endpoint training, utilizing both Microsoft and third-party technologies. Yeh's company assists clients in mandating and validating security training, often tying completion to employee KPIs for raises and bonuses. He explained that quantifying mandatory staff training and policy enforcement through phishing tests and other security reports can help clients reduce insurance costs and ensure policy enforceability in the event of a compromise, preventing insurance companies from denying claims.
According to Microsoft Security general manager Scott Woodgate, Project Perception demonstrates Microsoft's comprehensive security capabilities, extending beyond initial leaks that focused solely on vulnerability discovery. The system incorporates purpose-built AI models, including Microsoft’s MAI-Cyber-1, alongside organizational context, full-estate signals and sensors, an orchestration harness, and actuators that enable AI agents to execute decisions. Project Perception is designed to integrate with Microsoft’s generative AI-assisted Security Copilot chat interface and operates on a consumption-based, pay-as-you-go pricing model.