Users are required to comply with all applicable laws when using the model and are prohibited from generating harmful or deceptive content with it.
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 places affirmative compliance obligations on the deploying organization rather than on DeepSeek, meaning that legal responsibility for lawful and non-harmful use of the model's outputs rests with the user or deploying entity.
Interpretive note: The terms 'harmful' and 'deceptive' are not defined in the license, and the applicable legal standards vary significantly by jurisdiction and deployment context, creating interpretive uncertainty for deploying organizations.
Removal of explicit harmful use and deceptive content prohibitions delegates all use restrictions to general applicable law rather than contractual terms.
View full change record →Severity was increased from medium to high, indicating stricter enforcement priority.
View full change record →Severity was elevated from medium to high, indicating stricter enforcement expectations.
View full change record →Severity increased from medium to high, indicating DeepSeek now places greater emphasis on compliance obligations.
View full change record →Severity increased from medium to high, indicating stricter enforcement of acceptable use policies.
View full change record →Severity upgraded from medium to high, indicating increased importance of compliance enforcement.
View full change record →Organizations and developers deploying DeepSeek-R1 in products bear the contractual and legal responsibility for ensuring that the model's outputs comply with applicable laws and do not generate harmful or deceptive content; this obligation is not assumed by DeepSeek under the license.
How other platforms handle this
Medium may review your conduct and content for compliance with these Terms and our Rules, and reserves the right to remove any violating content.
players may report each other and submit info about their in-game activity and actions to help us determine if the reported behavior constitutes a violation of rules and policies that apply to player conduct
Zoom uses advanced tools to automatically scan certain types of content such as virtual backgrounds, profile images, incoming emails to Zoom's native email service from someone who is not a Zoom Email user, and files uploaded or exchanged through chat...
"You must comply with applicable laws and regulations in your use of the Model. You will not use the Model to generate content that is harmful, deceptive, or otherwise violates applicable laws.Excerpt from DeepSeek's Model License
1) REGULATORY LANDSCAPE: This provision interacts with a broad range of regulatory frameworks depending on the deployment context, including the EU AI Act (which imposes obligations on deployers of AI systems, particularly high-risk systems), FTC …
Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.
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 places affirmative compliance obligations on the deploying organization rather than on DeepSeek, meaning that legal responsibility for lawful and non-harmful use of the model's outputs rests with the user or deploying entity.
Organizations and developers deploying DeepSeek-R1 in products bear the contractual and legal responsibility for ensuring that the model's outputs comply with applicable laws and do not generate harmful or deceptive content; this obligation is not assumed by DeepSeek under the license.
ConductAtlas has identified this type of provision across 142 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by DeepSeek.