Microsoft readies Project Perception to challenge Anthropic in AI security

Microsoft is building a code scanning security tool that routes tasks across its own, OpenAI and Anthropic models to find, verify and fix vulnerabilities at lower cost than Anthropics Claude Mitos.