Anthropic's third misuse report says its 2025 Claude models sat well below the level that could help a sophisticated user with dangerous biology, and that for current models it can no longer make that assurance.
Perspective Coverage
5 publishers
- Builder
- Builder 28%
- Operator
- Operator 52%
- Investor
- Investor 20%
Reality
- Evidence50
- Adoption
- Insufficient
- Hype gap+20
- Incentives65
- Confidence55
The share buying model-serving platforms went from 4.5% in January to 6.1% in July, and two named buyers have now moved production work onto Chinese open weights while incumbent revenue has not yet moved.
Reality
- Evidence54
- Adoption38
- Hype gap+21
- Incentives66
- Confidence55
American labs still hold the frontier. They are losing the volume, and the pricing gap doing the damage is roughly 36x on cheap input tokens.
Reality
- Evidence30
- Adoption54
- Hype gap+30
- Incentives45
- Confidence33
The data-removal firm's 2026 ranking puts Gemini and Meta AI among the riskiest and Vibe, ChatGPT and Pi among the safest. It is a procurement input, but it grades policies, not contracts.
Reality
- Evidence32
- Adoption
- Insufficient
- Hype gap+34
- Incentives68
- Confidence44
Hugging Face now hosts 2.96 million public model repositories. The ones anyone actually pulls number in the tens of thousands, and the monthly parameter ceiling has been set in China all year.
Reality
- Evidence54
- Adoption71
- Hype gap+9
- Incentives68
- Confidence48