Microsoft introduced two new artificial intelligence-based security solutions on Monday aimed at helping organizations identify and mitigate cybersecurity risks more efficiently and cost-effectively.
The first tool, named MAI-Cyber-1-Flash, is a specialized AI model trained specifically to detect and address software vulnerabilities. Built on Microsoft’s MAI-Thinking-1 platform, the model leverages the company’s extensive experience from decades of vulnerability management and security incident response across its product portfolio. Microsoft indicated the model was developed in-house using high-quality training data and processes more than 1 trillion security signals daily while drawing insights from approximately 1.6 million customers.
The MAI-Cyber-1-Flash model has been integrated into MDASH, a multi-model system combining 100 security-focused AI agents designed to discover exploitable bugs. According to Microsoft, this combined system achieved a 96 percent score on the CyberGYM benchmark test, outperforming comparable offerings from Anthropic and Google. The company also stated that the updated MDASH costs approximately half as much as its previous version.
The second announcement involved Project Perception, a platform consisting of specialized AI agents that perform various security functions including vulnerability identification, risk assessment, and corrective action implementation. Microsoft said the platform intelligently selects which AI models to deploy based on task requirements and cost considerations, allowing organizations to use more expensive alternatives only when necessary for specialized tasks.
Microsoft framed these tools as responses to evolving cybersecurity challenges, noting that AI is increasing both the speed and scale of potential attacks while forcing security teams to manage increasingly complex digital environments with outdated approaches. Both tools are currently in preview mode.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI