Unverified50% confidenceFactExact time
The Preparedness Framework divides risk categories into four core domains: cybersecurity, chemical/biological/nuclear/radiological weapons (CBRN), model autonomy, and persuasion and deception
1
Sources
50%
Confidence
Long-term
Relevance
9/6/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedOpenAI released the Preparedness Framework, which categorizes risks into four levels — low, medium, high, and critical — and requires that only models with a risk assessment below 'high' may be deployed externally.73% similarUnverifiedAnthropic uses the RSP framework to evaluate dangerous capability thresholds in models, OpenAI uses the Preparedness Framework, and Google DeepMind uses the Frontier Safety Framework70% similarUnverified任何具备联网和代码执行能力的前沿模型在大规模部署前都应经过独立严格的安全评估，特别是针对网络攻击、供应链投毒等高危场景的专项测试68% similarVerifiedOrganizations such as METR (Model Evaluation and Threat Research) are developing standardized dangerous capability evaluations, including whether a model can provide biosynthesis guidance beyond publicly available literature and whether it can autonomously discover zero-day vulnerabilities.68% similarUnverifiedSeveral leading laboratories have published their respective Responsible Scaling Policies or Preparedness Frameworks, setting risk-level thresholds for model capabilities.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/867087API
curl https://kongchang.com/api/v1/knowledge/claims/867087MCP
get_claim(id=867087)