Claude Mythos 5 Leak: The AI Anthropic Says Is Too Powerful to Release
中文摘要
Claude Mythos 5 泄露，Anthropic 认为其过于强大而无法发布，可能发现数十年漏洞，引起专家担忧。
English Summary
Leaked Claude Mythos 5, deemed too powerful by Anthropic, found decades-old vulnerabilities, alarming experts and potentially shifting the AI race.
原文节选
The leaked model that reportedly found decades-old vulnerabilities, alarmed cybersecurity experts, and could change the AI race forever. Continue reading on GoPenAI »