The document discloses that OpenAI, Anthropic, and Google Cloud Vertex models may be used for background processing tasks such as summarization regardless of which model the user has selected for their primary AI interactions. Enterprise administrators can disable specific providers at the organizational level.
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
This provision establishes that user model selection does not fully constrain which inference providers receive code-derived data, as background tasks may route data to additional providers. Enterprise administrators have controls to disable specific providers, but individual users do not appear to have equivalent granular controls outside of zero-data retention mode.
The updated terms establish a default policy permitting Windsurf to use customer data for model training purposes to improve services. Previously, the company required explicit opt-in before any training use. Under the revised policy, data training occurs automatically for free and paid users unless they affirmatively opt out through the Data Controls settings page. Once disabled, the terms state your data will not be used for training and Zero Data Retention will be enabled with model providers. Enterprise customers operate under a different standard, requiring express prior written consent before any training use occurs.
View change record →The updated document establishes explicit commitments about how Windsurf protects data and manages security. The terms state that all data transmission is encrypted in transit and at rest, that access to production systems is restricted to a small number of employees or contractors based on business roles, and that production systems are monitored via logging, error handling, and monitoring dashboards. The document discloses that Windsurf obtained SOC 2 Type II certification as of March 2024 and that all employees and contractors are required to use multi-factor authentication and receive annual security training. These disclosures describe organizational practices rather than establishing new user-facing rights or obligations.
View change record →Under these terms, data including code context may be routed to OpenAI, Anthropic, and Google Cloud Vertex for background tasks such as summarization independent of the user's explicit model selection. Enterprise administrators can disable specific model providers for their organization, but individual user controls over this routing are not described beyond zero-data retention mode.
How other platforms handle this
we may share data between our affiliates for the safety and security of our users and may take necessary actions if we believe you have violated these Terms, including banning you from our Services and/or our affiliates' services...
Whenever we transfer personal data internationally, we use tools and transfer agreements to: make sure the data transfer complies with applicable law; and help to give your data the same level of protection as it has in the EU...
disclosure is required by a third-party to complete a transaction initiated by the user
"We may leverage OpenAI models independent of user selection for processing other tasks (e.g. for summarization). We may leverage Anthropic models independent of user selection for processing other tasks (e.g. for summarization). We may leverage these models independent of user selection for processing other tasks (e.g. for summarization).Excerpt from Windsurf's Security & Data Handling
1.
Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.
ConductAtlas detected a major restructuring of Meta’s privacy policy that removed detailed consumer rights disclosures and relocated them to separate documents.
Your genetic data may be transferred to a new owner as a business asset. Here is what the Terms of Service actually say and what you can do right now.
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 establishes that user model selection does not fully constrain which inference providers receive code-derived data, as background tasks may route data to additional providers. Enterprise administrators have controls to disable specific providers, but individual users do not appear to have equivalent granular controls outside of zero-data retention mode.
Under these terms, data including code context may be routed to OpenAI, Anthropic, and Google Cloud Vertex for background tasks such as summarization independent of the user's explicit model selection. Enterprise administrators can disable specific model providers for their organization, but individual user controls over this routing are not described beyond zero-data retention mode.
ConductAtlas has identified this type of provision across 288 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Windsurf.