**Anthropic将最强模型Claude Mythos 5用于网络安全防御**

_Anthropic puts its most powerful model Claude Mythos 5 to work for cyber defense_

> Anthropic已将其安全扫描工具Claude Security升级至Claude Mythos 5模型,可扫描代码库漏洞、提供严重性评级与CWE分类,并建议修复方案。同时,Anthropic正将Mythos 5集成至合作伙伴的安全产品,用于保护关键基础设施。

**来源信息**
- **来源**:The Decoder:AI News(RSS)
- **分类**:ai-products
- **发布时间**:2026-08-22 03:35(北京时间)
- **原文**:[打开原文](https://the-decoder.com/anthropic-puts-its-most-powerful-model-claude-mythos-5-to-work-for-cyber-defense)