Microsoft’s ‘Project Perception’ Could Challenge Anthropic’s Mythos in AI Security

Microsoft is reportedly developing Project Perception, an AI-driven security product that combines models from Anthropic, OpenAI, and Microsoft to scan enterprise IT environments for vulnerabilities while managing costs by routing queries to the most appropriate models; the report also covers Microsoft’s enterprise-focused AI strategy, competitive positioning against Anthropic/OpenAI, and export/vetting considerations but does not describe an active cyber incident.