Get the weekly 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. No account.
The policy states that content inputs and outputs from AI features, including text, images, video, audio, and precise location shared with My AI, are collected and used to improve machine learning and generative AI models. The policy acknowledges a privacy minimization intent but does not define specific limits on the categories of input data used for model training.
This analysis describes what Snapchat'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 conversational and multimedia content shared with Snap's AI features is used as training data for model improvement, which extends the processing purpose beyond the immediate user interaction. The scope of data categories covered, including precise location and audio, may require evaluation under GDPR consent and purpose limitation requirements as well as applicable U.S. state privacy laws governing sensitive data.
Interpretive note: The policy does not specify whether users can opt out of AI training data use specifically, and the operational scope of 'privacy minimization' is not defined, creating ambiguity about the limits of this processing purpose.
Under this clause, text, images, video, audio, and precise location data shared during AI feature interactions may be used to train and improve Snap's machine learning models, not only to generate responses in the moment. The agreement does not specify an opt-out mechanism specific to AI training data use within the main policy text.
Cross-platform context
See how other platforms handle AI Feature Data Use for Model Development and similar clauses.
Compare across platforms →Monitoring
Snapchat has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 20 platforms.
"content you share and generate with our AI Features (content, or "Inputs," including text, images, video, audio, precise location, and engagement, used to generate content and responses, or "Outputs"). ... our algorithms and machine learning models take into account the conversations Snapchatters are having with My AI to improve the responses from My AI. Your information can help us decide what kind of improvements we should make, but we are always focused on privacy — and we don't want to use more of your personal information than necessary to develop our features and models.Excerpt from Snapchat's Privacy Policy
1) REGULATORY LANDSCAPE: This provision implicates GDPR purpose limitation and data minimization principles, as well as consent requirements where special category data may be inferred from AI inputs. The FTC Act's unfair or deceptive practices standard is relevant to the extent that the scope of AI training use materially exceeds user expectations based on disclosed purposes. U.S. state privacy laws in California and other states with enacted frameworks may classify certain AI-derived inferences as sensitive data subject to additional consent requirements. Relevant enforcement authorities include EU member state data protection authorities, the UK ICO, the FTC, and California Privacy Protection Agency (CPPA). 2) GOVERNANCE EXPOSURE: High. The collection and use of conversational AI inputs including location and audio for model training represents a broad processing purpose that intersects with data minimization obligations under GDPR and equivalent frameworks. The policy's privacy minimization language is aspirational rather than operationally binding, creating potential exposure if the scope of training data use is broader than what users reasonably anticipate from the disclosed purpose. 3) JURISDICTION FLAGS: EU and EEA users face heightened exposure given GDPR purpose limitation and consent requirements for secondary processing of personal data. California users may have rights under CCPA to opt out of certain uses of personal information for automated decision-making or profiling. Users sharing sensitive personal information through AI features, such as health-related content or location patterns, face additional exposure in jurisdictions with sensitive data protections. 4) CONTRACT AND VENDOR IMPLICATIONS: Organizations deploying Snapchat in enterprise or educational contexts should assess whether employee or student use of My AI features results in organizational data being processed for model training. Snap's Terms of Service and this policy do not appear to include enterprise data processing agreements limiting AI training use, which may be relevant for procurement teams. 5) COMPLIANCE CONSIDERATIONS: Compliance teams should audit whether in-app disclosures at the point of AI feature use adequately inform users that inputs will be used for model training, and whether those disclosures satisfy jurisdiction-specific informed consent standards. A data mapping update should confirm which categories of AI input data are retained, for how long, and under what legal basis for each jurisdiction. Where GDPR applies, the legal basis for AI training use, whether legitimate interest or consent, should be documented and reviewed for adequacy.
Regulatory citations, enforcement risk, and due diligence action items.
Provision-level monitoring, governance timelines, and regulatory mapping built from archived source documents and historical version tracking.
This provision establishes that conversational and multimedia content shared with Snap's AI features is used as training data for model improvement, which extends the processing purpose beyond the immediate user interaction. The scope of data categories covered, including precise location and audio, may require evaluation under GDPR consent and purpose limitation requirements as well as applicable U.S. state privacy laws …
Under this clause, text, images, video, audio, and precise location data shared during AI feature interactions may be used to train and improve Snap's machine learning models, not only to generate responses in the moment. The agreement does not specify an opt-out mechanism specific to AI training data use within the main policy text.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Snapchat.