Unverified50% confidenceFactExact time
The EU AI Act uses a training compute threshold of 10^25 FLOP as one of the benchmarks for determining frontier models.
1
Sources
50%
Confidence
Long-term
Relevance
10/10/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedThe EU AI Act categorizes generative AI under General Purpose AI (GPAI), requiring public disclosure of training data summaries, compliance with copyright regulations, and additional systemic risk assessments for models exceeding a specific FLOPs threshold.71% similarVerified欧盟《AI法案》要求AI公司披露训练数据来源70% similarUnverified拜登时代的AI行政令曾以10²⁶次浮点运算（FLOP）作为训练算力监管红线69% similarUnverifiedA project called "Show HN: How Stale Is Your AI?" systematically compiles the release dates and training data cutoff dates of 20 mainstream AI models.69% similarUnverifiedClaude believes the root cause of AI redundancy lies in the training corpus itself, as training data for analytical tasks is inherently structured as multi-point enumeration68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/994840API
curl https://kongchang.com/api/v1/knowledge/claims/994840MCP
get_claim(id=994840)