Anthropic 切断其内部评估与互联网的连接
Anthropic is cutting off its internal evaluations from the internet
After a recent spate of high-profile incidents in which AI agents escaped containment, Anthropic is cutting off internet access for all internal evaluations. In a report Friday, the company detailed "unintended model actions," including submitting a false tip regarding an unsolved murder, that led to the decision. Although the impact of these behaviors was minimal […]
来源 theverge.com
同题报道 · 6 篇
- Anthropic 切断其内部评估与互联网的连接 Anthropic is cutting off its internal evaluations from the internet theverge.com ·
- Anthropic 切断其内部评估与互联网的连接 Anthropic is cutting off its internal evaluations from the internet The Verge ·
- Anthropic 在 Claude AI 模型绕过安全措施后限制互联网访问 Anthropic Restricts Internet Access After Claude AI Models Bypass Safeguards indica News ·
- Anthropic 在 Claude 绕过安全措施后限制互联网访问 Anthropic restricts internet access after Claude bypasses safeguards Awaz The Voice ·
- Anthropic 在 Claude AI 绕过安全措施后限制其互联网访问 Anthropic restricts Claude AI internet access after bypassing safeguards Asianet Newsable ·
- Anthropic 在 Claude 绕过安全措施后限制内部 AI 评估的互联网访问 Anthropic restricts internet access in internal AI evaluations after Claude bypasses safeguards BusinessLine · · 首发