This clause prohibits using DeepSeek-R1 or any outputs it generates to train, improve, or fine-tune any competing AI language model.
This analysis describes what DeepSeek's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
This provision directly limits how AI developers and researchers can use the model's outputs, specifically barring a common practice of using outputs from one model to improve or train another, which has significant implications for organizations operating AI model development programs.
Interpretive note: The enforceability of anti-distillation restrictions as contractual terms versus legally permissible uses under copyright fair use or text and data mining exceptions varies by jurisdiction and remains legally unsettled.
Removal of the anti-distillation restriction eliminates a major constraint on model improvement and competitive development, significantly liberalizing permitted uses.
View full change record →For individual developers and organizations that use DeepSeek-R1 outputs as training data or as inputs for improving other AI models, this clause represents a contractual restriction that, if violated, could constitute a license breach. The restriction applies to outputs generated by the model, not just the model weights themselves.
How other platforms handle this
Microsoft reserves the right to restrict the export of data that may compromise the security of the services or Microsoft's intellectual property.
Datadog reserves the right, but does not assume the obligation, to investigate any violation of these Terms or misuse of the Site.
Pharmaceutical products
"You will not use the Model or any output of the Model to improve any other large language model (excluding DeepSeek-R1 and its derivatives).Excerpt from DeepSeek's Model License
1) REGULATORY LANDSCAPE: This clause operates primarily under intellectual property and contract law rather than consumer data protection frameworks.
Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.
Buried in Robinhood's customer agreement is broad authority to close your positions, suspend your account, and force arbitration. Here is what it actually says.
Stripe's terms authorize fund reserves, payout withholding, and account termination. Here is what the agreement states and what business owners should review.
Get the research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.
This provision directly limits how AI developers and researchers can use the model's outputs, specifically barring a common practice of using outputs from one model to improve or train another, which has significant implications for organizations operating AI model development programs.
For individual developers and organizations that use DeepSeek-R1 outputs as training data or as inputs for improving other AI models, this clause represents a contractual restriction that, if violated, could constitute a license breach. The restriction applies to outputs generated by the model, not just the model weights themselves.
ConductAtlas has identified this type of provision across 263 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by DeepSeek.