Unverified50% confidenceFactExact time
The U.S. 2023 executive order requires safety evaluations for models trained at large scale, but primarily constrains the training parties and lacks effective jurisdiction over downstream users' modification behaviors.
1
Sources
50%
Confidence
Long-term
Relevance
9/5/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedExecutive Order 14110要求超过10^26次FLOPs训练量的模型需强制向商务部NIST、国土安全部等机构提交安全评估报告75% similarUnverified美国AI监管以模型训练算力阈值（以FLOP为单位）作为触发强制报备义务的量化指标69% similarUnverified美国BIS于2023年发布的芯片出口管制规则已将特定算力门槛以上的模型训练纳入管制考量69% similarUnverifiedExecutive Order 14110以训练所用浮点运算量（FLOPs）作为触发审查义务的量化阈值，超过10²⁶次FLOPs训练量的模型即需强制报备69% similarUnverifiedThe training-time scaling law was proposed by Kaplan et al. in 2020, holding that model capability is primarily determined by parameter count, training data volume, and training compute.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/861887API
curl https://kongchang.com/api/v1/knowledge/claims/861887MCP
get_claim(id=861887)