Unverified50% confidenceFactExact time
U.S. copyright law has no clear precedent on whether model weights constitute derivative works, and the EU AI Act does not directly regulate model archiving behavior either.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/20/2026
First Seen
Valid until: 12/19/2026
Sources
Related Entities
Related Claims
Unverified模型权重的版权归属在全球范围内仍处于法律真空地带，对于模型权重本身是否构成受保护的创作性表达尚无权威判决80% similarVerifiedThe U.S. Copyright Office has explicitly stated that works purely generated by AI are not protected by copyright, but works in which humans have made substantial modifications to AI output may receive limited protection.76% similarUnverifiedThe U.S. Copyright Office's 2023 AI copyright guidelines have not yet clearly defined the copyright ownership of model outputs used as training data.76% similarUnverifiedThe EU AI Act has not fully clarified the boundary between AI training and copyright.75% similarVerifiedThe U.S. Copyright Office's 2023 report "Artificial Intelligence and Copyright" states that purely AI-generated content does not qualify for copyright protection (establishing precedent in Thaler v. Vidal).71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/937205API
curl https://kongchang.com/api/v1/knowledge/claims/937205MCP
get_claim(id=937205)