OpenAI Terms of Use

This is OpenAI's Terms of Use — the legal agreement you accept when using ChatGPT, Sora, Codex, or any other OpenAI product. The most important thing to know is that by default, OpenAI may use your conversations and inputs to train its AI models, though you can opt out of this data use through your account settings. If you are in the EU, a separate set of terms applies to you, and if you are a business or developer using the API, yet another agreement governs your relationship with OpenAI.

This document constitutes OpenAI's Terms of Use governing consumer and developer access to OpenAI's AI services including ChatGPT, Sora, Codex, and the OpenAI API, establishing a contractual relationship under applicable law with OpenAI, Inc. as the contracting entity. The terms impose significant obligations on users including compliance with OpenAI's Usage Policies, restrictions on reverse engineering or misuse of AI outputs, age verification requirements (minimum age 13, or 18 for certain services), and grant OpenAI a broad license to use user-submitted content for service improvement and model training purposes. Notable deviations from industry standard include OpenAI's assertion of the right to use user inputs and outputs to train AI models (subject to opt-out mechanisms), a tiered terms structure that routes API and business users to separate agreements, and differentiated terms for EU users via a separate EU Terms of Use document. The document engages GDPR (for EU/EEA users routed to separate EU terms), CCPA (for California residents), COPPA (age restrictions, 13-year minimum), the EU AI Act (AI-generated content obligations), and FTC Act Section 5 (unfair or deceptive practices); compliance teams should note that the document's reference to separate EU Terms and Business Terms creates a multi-tier legal framework that requires independent review of each instrument for full regulatory exposure mapping.