白宫恢复英国 AISI 前沿模型访问
英国 AI 安全机构将重新获得多家公司新模型的早期访问权。
为什么重要
这关系到前沿模型发布前的跨国安全测试与监管协作。
阅读原文the-decoder.com
AI 解读（根据原文生成）
The Decoder 援引 The Daily Telegraph 报道称，白宫撤销了阻止英国 AI Security Institute（AISI）访问新 AI 模型的决定。AISI 负责在 AI 系统发布前测试安全风险，将再次获得 Anthropic、Google DeepMind 和 OpenAI 模型的早期访问。此前特朗普政府曾要求这些公司停止与 AISI 共享模型，Anthropic 随后限制了 Claude Mythos 5.1 的访问。报道还提到，AI Minister Kanishka Narayan 计划宣布新的 AI 安全措施。
要点（AI 据原文整理，数字均已与原文核对）
- The White House has reversed its decision to block the UK's AI Security Institute (AISI) from accessing new AI models.
- AISI tests AI systems for safety risks before they ship.
- AISI will now once again get early access to models from Anthropic, Google DeepMind, and OpenAI.
- Back in September, the Trump administration had told the companies to stop sharing their models with AISI.
- Anthropic then restricted access to its new Claude Mythos 5.1 model.
- Narayan plans to announce new AI safety measures, including stricter reporting requirements for AI-related security incidents involving critical infrastructure.
