This analysis describes what Windsurf'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
The provision establishes a restricted use framework for user-generated autocomplete data, distinguishing between permissible discriminative applications and prohibited generative applications. This operational distinction defines the scope of data utilization for model improvement while preserving user control through opt-out mechanisms.
The updated terms indicate that Windsurf is now operating as the Cognition Platform under Cognition AI, Inc., replacing the prior Exafunction, Inc. structure. The revised terms state that prior terms continue to govern use for 30 days from the posting date (July 1, 2026), and that continued access after that period constitutes acceptance of the updated terms. Users who do not agree with the new terms are instructed to stop using or accessing the Services. The specific substantive changes to user rights, data collection, fees, or service functionality are not detailed in the provided change summary.
View change record →Users' Autocomplete User Content may be used for discriminative model training by default unless the user modifies settings to opt out. If a user does not opt out, anonymized autocomplete data will be retained and used for training discriminative models that rank code generations.
How other platforms handle this
Data publicly available on the Internet. Our artificial intelligence models are trained on data that is publicly available on the Internet by third parties, which may contain personal data, even if we use good practices to filter out such personal data. [...] Training Datasets. In some cases, we acc...
To improve the quality of our services, we analyse texts submitted for translation. We ensure that this analysis cannot be traced back to individual users by anonymising the data before analysis. DeepL Pro subscribers' texts are not used to train our machine translation systems.
Only models with a post-mitigation score of "medium" or below can be deployed. Only models with a post-mitigation score of "high" or below can be developed further.
Monitoring
Windsurf has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 25 platforms.
"We may use your Autocomplete User Content to improve our discriminative machine learning models, which are models that rank or assign scores to code generations in order to understand the boundaries between different sets of code. We will never use your Autocomplete User Content to improve generative machine learning models, which are models that are able to generate code directly based on studying existing code generations, for Autocomplete or other services. Any Autocomplete User Content used for training our discriminative machine learning models is anonymized, such that any personally identifiable information is removed. To opt out of having your Autocomplete User Content used for such purpose, you may change the code sharing options in the User Settings pane of the user's profile page. Please note that if you opt out, your Autocomplete User Content will be sent to our servers so that we are able to provide you with the Services, but we will not retain your Autocomplete User Content on our servers for training our discriminative machine learning models.— Excerpt from Windsurf's Windsurf Terms of Service
How Meta, TikTok, and Supabase restructured governance language across documents, jurisdictions, and consent frameworks through incremental document updates.
How 10 AI platforms describe the use of user data for model training, improvement, and development, based on archived governance provisions.
Compliance Governance Intelligence
Need to monitor specific governance provisions?
Compliance includes provision-level monitoring, governance timelines, regulatory mapping, and audit-ready analysis.
Built from archived source documents, structured governance mappings, and historical version tracking.
The provision establishes a restricted use framework for user-generated autocomplete data, distinguishing between permissible discriminative applications and prohibited generative applications. This operational distinction defines the scope of data utilization for model improvement while preserving user control through opt-out mechanisms.
Users' Autocomplete User Content may be used for discriminative model training by default unless the user modifies settings to opt out. If a user does not opt out, anonymized autocomplete data will be retained and used for training discriminative models that rank code generations.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Windsurf.