[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-anthropic-mythos-project-glasswing-withhold":3,"topics-all":33,"news-related-700acd0e-37e1-4ea0-94c2-ded415f2de95":52},{"id":4,"title":5,"summary":6,"content":6,"original_url":7,"source_id":8,"tags":9,"translations":20,"news_slug":26,"published_at":27,"created_at":28,"modified_at":29,"is_published":30,"publish_type":31,"image_url":13,"view_count":32},"700acd0e-37e1-4ea0-94c2-ded415f2de95","Anthropic 首次主动扣留模型：Claude Mythos 安全能力过强引发行业担忧","Anthropic首次公开扣留旗舰模型。5月，Anthropic宣布通过Project Glasswing项目，向约50家科技巨头（微软、谷歌、苹果、NVIDIA、摩根大通等）提供Claude Mythos Preview的受限访问，而非公开发布。这款模型在网络安全测试中展现出惊人的攻防双向能力——它不仅能发现开源项目中的23019个高危和严重漏洞（90.6%经独立采样确认真实），还能将多个漏洞串联利用，自主写出可工作的攻击代码。这让Anthropic决定暂时不让所有人都有权访问。\n\n上一次头部AI公司因安全顾虑主动留一手，还是2019年OpenAI扣留GPT-2。Mythos是迄今为止技术力量最强的一次例外。\n\n为什么这事值得关注？因为它触及了AI行业最核心的张力：模型能力越强，潜在的武器化风险就越高，但同时防御方也能用同样的工具来加固系统。Anthropic选择了一条中间道路——让防御者先用起来，同时向政府做了全面汇报。但这并不意味着问题解决了。Anthropic内部研究人员坦言，在模型足够强大到能自动化高级持续性渗透的当下，如何准备一个AI网络战成为现实的世界，目前还没有完整答案。\n\n这给行业敲了一记警钟：当模型的offensive capability开始与最强的国家级黑客工具相当，单纯靠发布后打补丁的策略还管用吗？Mythos的故事说明，AI安全不能只靠对齐研究，还需要安全研究员、政策制定者和模型厂商三方协同，在模型出门之前就把路修好。","https:\u002F\u002Fwww.nbcnews.com\u002Ftech\u002Fsecurity\u002Fanthropic-project-glasswing-mythos-preview-claude-gets-limited-release-rcna267234","226bcb3d-18b8-4bb0-a999-4e82ec13f5fd",[10,14,17],{"id":11,"name":12,"slug":12,"description":13,"color":13},"e676a5cf-1f24-472f-a765-86fa21a1bc3c","ai-model",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"1fcfaaf2-67de-43d3-9e35-5784852fec60","ai-safety",{"id":18,"name":19,"slug":19,"description":13,"color":13},"23544f6a-eea1-4f05-aa8d-749ca862d5d2","anthropic",[21],{"id":22,"lang":23,"title":24,"summary":25,"content":13},"15d8984f-d091-4507-b6ea-cd228acf8638","en","Anthropic withholds a model for being too capable","Anthropic announced on June 2 that it is withholding the full release of Claude Mythos, citing concerns about the model's safety capabilities being too strong. This is the first time Anthropic has actively limited a model's release due to safety considerations, marking a shift from \"release fast, fix later\" to \"release with caution.\"","anthropic-mythos-project-glasswing-withhold","2026-06-02T08:10:00Z","2026-06-02T16:06:32.769538Z","2026-08-19T02:08:40.142862Z",true,"agent",124,[34,43],{"slug":35,"tag_slug":35,"title_zh":36,"title_en":37,"intro_zh":38,"intro_en":39,"id":40,"is_active":30,"created_at":41,"modified_at":42},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":44,"tag_slug":44,"title_zh":45,"title_en":46,"intro_zh":47,"intro_en":48,"id":49,"is_active":30,"created_at":50,"modified_at":51},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":53},[54,59,64,69,74,79],{"id":55,"title":56,"news_slug":57,"published_at":58},"1c0ab6ec-1647-4820-a61c-7eef232075b9","Anthropic 发现训练数据配比新方法：让 AI 对齐从「事后再修」走向「源头预防」","anthropic-training-data-ratio-alignment-zero-blackmail","2026-05-11T01:00:00+00:00",{"id":60,"title":61,"news_slug":62,"published_at":63},"fa4c43df-5d62-464b-b4b0-3a6854000644","AI 攻破 OpenAI 内网全程复盘:靠 Anthropic 模型当武器,72 小时打开自家后门","hacktron-claude-openai-monorepo-rce-72h","2026-09-21T03:00:00+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"95e9bb62-0bd3-4c2f-913a-302ba5e2ace8","Anthropic 9 月报告把蒸馏战摆上台面:151 亿次阿里请求、解放军流量走 Moonshot","anthropic-distillation-report-china-200m-claude","2026-09-18T03:00:00+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"d9a24a72-c5b8-4117-b8c2-a9981180c8bd","AI 三巨头罕见同框：马斯克、Altman 站队 Amodei 喊停前沿研发","amodei-musk-altman-pace-ai-frontier","2026-09-14T02:00:00+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"f3d17d45-e1a8-4a1b-9449-6813aff06e49","Anthropic 让 Claude 自己修对齐:10 类失败全部见效,还超过人类研究员","claude-automated-alignment-researchers","2026-08-29T13:05:00+00:00",{"id":80,"title":81,"news_slug":82,"published_at":83},"7dec6918-b6cb-4b85-a6bf-88d1abc332d0","加密推理块漏洞让 Anthropic\u002FOpenAI\u002FGoogle 的思维链全部裸奔","stealing-reasoning-traces-llm-apis","2026-08-21T10:00:00+00:00"]