360创始人周鸿祎发布AI安全工具对标Mythos,承认落后20%-30%

Chinese cybersecurity firm builds AI tools to rival Mythos and frames the race as cyber-nuclear deterrence

精选理由

360推出了对抗Anthropic Mythos的AI安全工具,其中一个已经发现了3400多个漏洞,但周鸿祎实话实说咱们的模型比西方差两成。

AI 摘要

360创始人周鸿祎发布两款AI安全工具,旨在与Anthropic的Mythos竞争。其中一款工具已标记3,432个漏洞。周鸿祎承认中国模型在性能上落后西方20%至30%。他将Mythos比作“网络核武器”,并呼吁中国建立自己的战略威慑能力。

AI 翻译 · 中文

360创始人周鸿祎发布两款AI安全工具,旨在与Anthropic的Mythos竞争。其中一款工具已标记3,432个漏洞。周鸿祎承认中国模型在性能上落后西方20%至30%。他将Mythos比作“网络核武器”,并呼吁中国建立自己的战略威慑能力。

Decoder360 founder Zhou Hongyi presents two AI security tools designed to compete with Anthropic's Mythos. One has already flagged 3,432 vulnerabilities. Zhou admits Chinese models trail Western ones by 20 to 30 percent, but co