Unverified50% confidenceFactExact time
Jan Leike在离职时公开表示OpenAI的安全文化和流程已经让位于闪亮的产品
1
Sources
50%
Confidence
Long-term
Relevance
8/21/2026
First Seen
Sources
Related Claims
UnverifiedJan Leike, who was responsible for the superalignment team, joined Anthropic after publicly criticizing OpenAI for prioritizing products over safety73% similarUnverifiedOpenAI is advancing a Trusted Access Program for cybersecurity.64% similarUnverifiedValuations of companies such as OpenAI and Anthropic have repeatedly hit new highs.63% similarUnverifiedOpenAI的InstructGPT、Anthropic的Constitutional AI以及Meta的Llama系列都采用了类似的安全对齐流程63% similarUnverifiedOpenAI有Preparedness Framework，对模型在网络安全、CBRN威胁、说服力和自主行为四个维度进行分级评估61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/782659API
curl https://kongchang.com/api/v1/knowledge/claims/782659MCP
get_claim(id=782659)