Why did Anthropic deny the UK AI Security Institute (AISI)—which has a £66 million annual budget, more than 100 technical specialists, has eThe dispute highlights the tension between controlled access to high-capability models and independent safety evaluation.
AI 提示
Create a landscape editorial hero image for this Studio Global article: Why did Anthropic deny the UK AI Security Institute (AISI)—which has a £66 million annual budget, more than 100 technical specialists, has e. Article summary: Anthropic has not publicly stated why it withheld Mythos 5.1 from AISI, so a definitive motive is unknown. The evidenced fact is narrower: the institute did not receive pre-release access while a small, vetted US group d. Topic tags: general, general web, user generated, government, documentation. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text,
openai.com
報道指,Anthropic 並無向英國 AI Security Institute(AISI,英國政府轄下、專責研究先進 AI 風險嘅機構)提交 Claude Mythos 5.1 作發佈前測試。呢次據稱係 Anthropic 首次喺前沿模型評估中未有讓 AISI 參與。現有資料只證實「冇畀存取權」呢個決定,並無證實背後動機:Anthropic 至今未公開交代點解排除 AISI。 34