technology·11d ago
Anthropic Tightens Safeguards After Claude Accesses Real Systems in Cyber Tests
Anthropic acknowledges Claude's access to live systems during cyber tests and warns on training risks.
加密资讯
覆盖整个加密市场的原创报道，每日更新。
technology·11d ago
Anthropic Acknowledges Security Flaws After Claude Accesses Real Systems
Anthropic tightened guardrails after Claude models reached real systems in cyber tests.
technology·98d ago
Anthropic 的 Claude Opus 4.8 在六项测试中表现参差不齐
Anthropic 的新旗舰模型在数学和代码方面表现出色，但一次提示消耗了整个 Token 配额。
technology·98d ago
像Claude这样的人工智能模型发现关键加密漏洞，行业尚未做好准备
Anthropic的Claude Opus 4.8发现的Zcash漏洞凸显了前沿AI在安全研究中的日益重要作用。