METAL

Anthropic

前沿实验室美国

打造Claude的AI公司。由前OpenAI员工于2021年创立,以「安全AI」为招牌。2026年8月发布的Claude Fable 5是一款将生物相关风险拦截率降至85%的模型。同月,Anthropic组建了自有的AI芯片设计团队,开始具备硬件开发能力。2026年8月,Anthropic披露Claude在评估过程中未经授权访问了三家真实企业的系统。

官方网站 ↗

了解 Claude 的含义与使用方法

当前排名 (1M)

2位

变化
上期
1

最近 30 天排名走势

2026.08.17 – 2026.09.15 · 最高 1位 · 最低 2位 · 当前 2位

Anthropic · 相关文章

AI

Claude Fable 5, 生物安全误判率降低85%

Anthropic表示已重新设计Fable 5的生物安全分类器,将因误判(false positive)导致的回退发生率降低约85%。病毒学、毒理学、分子设计等具有双重用途风险的领域仍维持拦截。

作者 김현국

00

Anthropic · 最新发布

  1. 80,939앤스로픽, 클로드 악용 8개월치 다 공개했다기사2026-09-10
  2. 37,586Anthropic’s Economics team is sharing a new model of how AI might affect economic growth, jobs, wages, and more by 2030.2026-09-09
  3. 15,703클로드, 페르마의 마지막 정리 첫 컴퓨터 검증 증명 완성기사영상2026-09-04
  4. 718Claude Fable 5.1 and Claude Mythos 5.1영상2026-09-01
  5. 15,402Today, we're kicking off the first phase of the research preview for Model Hardware Standard (MHS): a new standard for AI agents to safely operate physical equipment in scientific research and advanced manufacturing.영상2026-08-27
  6. 11,403앤스로픽, 클로드 사고를 미정렬로 고쳐 말했다기사2026-09-09
  7. 4,256앤스로픽, 클로드 무단접속 사고 후 훈련까지 중단기사2026-08-31
  8. 3,481New Fellows Research: Can Claude autonomously align other AIs?2026-08-28
  9. 222What will our economic future look like?2026-09-09
  10. 2,845For the first time, we’ve given external researchers a way to study AI’s impacts using real, privacy-preserved Claude usage data. To date, this work has only been possible within AI labs. We can’t tell the whole story alone, so we opened up our tools.2026-08-26
  11. 202클로드 코드, 커밋마다 세션 링크 자동 삽입해 논란기사2026-08-30
  12. 309앤스로픽, 보상 해킹만 가르쳤더니 사이버공격까지 배웠다기사2026-09-01
  13. 24Anthropic/enabling-independent-research · dataset2026-08-26
  14. –Detecting and countering misuse of AI: September 20262026-09-10
  15. –Meet Claude Fable 5.1영상2026-09-01
  16. –Introducing Claude Fable 5.1영상2026-09-01
  17. –앤스로픽, 기업 데이터 저장을 고객 클라우드로 넘긴다기사2026-09-01
  18. –Improving our alignment and security efforts2026-08-31
  19. –앤스로픽, 클로드에 AI 정렬 연구 맡기니 사람 앞질렀다기사2026-08-28
  20. –Model Hardware Standard: AI operating physical equipment영상2026-08-28
  21. –We're building a way for AI models to connect to any device and run real experiments.영상2026-08-27
  22. –AI models can now help run physical science experiments영상2026-08-27
  23. –ORDER ON CROSS MOTIONS FOR SUMMARY JUDGMENT2026-08-27