Microsoft reportedly building a Mythos rival
Multiple outlets, citing The Information, report Microsoft is building an AI security tool, Project Perception, to compete with Anthropic's Mythos.
The Information reports, and several outlets including TechRepublic and Neowin have followed up citing that report, that Microsoft is preparing an AI-driven cybersecurity tool internally codenamed Project Perception, aimed as soon as this month at competing with Anthropic’s Mythos vulnerability-finding model. As described, it would scan a company’s code, cloud infrastructure and endpoints, identify exploitable weaknesses and propose fixes, routing tasks across a mix of Anthropic, OpenAI and Microsoft’s own models with cost as its main selling point against Mythos. Microsoft has not confirmed the product publicly, and no outlet has cited an on-record Microsoft source; this remains an unconfirmed report pending an official announcement.