Microsoft unveils cost-saving AI model for cybersecurity, claims edge over Anthropic
Microsoft announced its first proprietary AI model for cybersecurity, MAI-Cyber-1-Flash, designed to identify vulnerabilities in source code. When paired with OpenAI's GPT-5.4, the company claims it outperforms Anthropic's Mythos 5, Google's 3.5 Flash Cyber, and OpenAI's GPT-5.5 Cyber on the CyberGym benchmark at 50% of the cost. The model will be integrated into Project Perception, a suite of AI agents for vulnerability discovery and remediation, entering public preview on August 3. The announcement marks Microsoft's first major cybersecurity push since Hayete Gallot returned to lead the unit in February. Microsoft AI CEO Mustafa Suleyman touted world-leading performance at reduced cost. The move comes as Microsoft shares have fallen 19% in 2026 amid investor concerns about its reliance on OpenAI, though analysts recommend buying. The company is also developing in-house models for GitHub Copilot and Excel to improve spending efficiency.
Editorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page itself is projected from evidence records.
- Current automated evidence projection