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.
This provision explicitly prohibits Perplexity and any third parties it engages from using Customer-submitted inputs or generated outputs to train, retrain, fine-tune, or otherwise improve AI models.
This analysis describes what Perplexity AI'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 a contractual commitment that distinguishes API Customer Content from data Perplexity may otherwise use for model improvement purposes. The commitment extends to third-party model providers authorized by Perplexity, creating a downstream obligation that compliance teams should verify is operationally implemented and contractually backed in Perplexity's vendor agreements.
Under this clause, Perplexity commits that neither it nor authorized third parties will use Customer inputs or API-generated outputs to train or improve AI models. This provision applies to the API service only and does not govern data practices for Perplexity's consumer products.
Cross-platform context
See how other platforms handle No Model Training on Customer Content and similar clauses.
Compare across platforms →Monitoring
Perplexity AI has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 25 platforms.
"Perplexity shall not use (or authorize third parties to use) Customer Content to train, retrain, fine-tune or otherwise improve any generative artificial intelligence models.Excerpt from Perplexity AI's Perplexity API Terms of Service
1) REGULATORY LANDSCAPE: This provision engages GDPR Article 5 (purpose limitation) and CCPA provisions governing secondary use of personal data, as well as emerging AI governance frameworks including the EU AI Act, which imposes obligations on providers of general-purpose AI models regarding training data. The FTC's enforcement posture on unfair or deceptive AI data practices is also relevant. Where Customer Content contains personal data, use for model training without an appropriate legal basis would implicate GDPR enforcement by EU supervisory authorities. 2) GOVERNANCE EXPOSURE: Medium. The provision creates a contractual obligation but does not specify audit rights, verification mechanisms, or remediation procedures if a violation occurs. The absence of an audit right means Customer has limited ability to independently verify compliance with this commitment. 3) JURISDICTION FLAGS: EU and EEA customers face heightened exposure if this commitment is not operationally implemented, given GDPR's purpose limitation and data minimization principles. California customers may evaluate this commitment against CCPA's restrictions on secondary use of personal information. 4) CONTRACT AND VENDOR IMPLICATIONS: Procurement teams should assess whether Perplexity's contractual commitment to third-party model providers mirrors the no-training obligation asserted here. The agreement does not provide Customer with audit rights or breach notification specific to this provision, which may require negotiation for enterprise deployments. 5) COMPLIANCE CONSIDERATIONS: Legal teams should document this commitment in vendor assessments and data processing records. Where Customer Content contains personal data, teams should evaluate whether the DPA adequately addresses this commitment and whether data mapping records reflect the restriction on model training use.
This provision establishes a contractual commitment that distinguishes API Customer Content from data Perplexity may otherwise use for model improvement purposes. The commitment extends to third-party model providers authorized by Perplexity, creating a downstream obligation that compliance teams should verify is operationally implemented and contractually backed in Perplexity's vendor agreements.
Under this clause, Perplexity commits that neither it nor authorized third parties will use Customer inputs or API-generated outputs to train or improve AI models. This provision applies to the API service only and does not govern data practices for Perplexity's consumer products.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Perplexity AI.