# Microsoft preps Project Perception to rival Anthropic's Mythos

> Microsoft is building a multi-model AI security platform to undercut Anthropic's Mythos.

*The cybersecurity platform mixes AI models from Microsoft, OpenAI and Anthropic to hunt and fix vulnerabilities more cheaply.*

By Behzad Hosseini · FeaturedDaily
Canonical: https://featureddaily.com/news/microsoft-preps-project-perception-to-rival-anthropic-s-mythos

**What happened:** Microsoft is building Project Perception, an AI cybersecurity platform that uses multiple AI models to detect and repair software vulnerabilities, The Information reported, as summarised by [TechRepublic](https://www.techrepublic.com/article/news-microsoft-project-perception-ai-security-tool/) on 17 July.

**The details:** A model router assigns each task to the best-suited model. Cheaper models handle inventory, log parsing, codebase navigation and triage. More capable models take on exploit-path reasoning and writing fixes. The platform draws on models from Microsoft, OpenAI and Anthropic.

**The numbers:** Microsoft's earlier MDASH system scored 88.45% on the CyberGym benchmark in May 2026, ahead of Anthropic's Claude Mythos 5 (83.1%) and GPT-5.5 (81.8%). Mythos is listed at $10 per million input tokens and $50 per million output tokens.

**The rivals:** Project Perception is positioned directly against Mythos, which Microsoft says is strong at vulnerability-hunting but expensive and limited to vetted security partners. Microsoft wants to match that analysis at lower cost.

**Why it matters:** AI models are now finding real software flaws faster than humans. Whoever sells that capability to defenders cheaply and broadly gains an edge in enterprise security.

**Who's affected:** Microsoft already sells Defender, Sentinel, Entra and Security Copilot to enterprise customers. Project Perception would be its first major security push since a leadership shake-up in February.

**What's next:** The platform was expected to be unveiled at a company event later in July, with a public preview to follow, according to [CNBC](https://www.cnbc.com/2026/07/27/microsoft-touts-cost-saving-ai-model-for-cybersecurity.html).

## Key takeaways

- Project Perception routes tasks across Microsoft, OpenAI and Anthropic models.
- It follows MDASH, which beat Mythos and GPT-5.5 on CyberGym.
- Mythos costs $10/$50 per million tokens; Microsoft aims to undercut that.

## Sources

- [Microsoft's 'Project Perception' Could Challenge Anthropic's Mythos in AI Security](https://www.techrepublic.com/article/news-microsoft-project-perception-ai-security-tool/) — TechRepublic, 2026-07-17
- [Microsoft touts cost-saving AI model for cybersecurity](https://www.cnbc.com/2026/07/27/microsoft-touts-cost-saving-ai-model-for-cybersecurity.html) — CNBC, 2026-07-27