What each tool's own terms say about your prompts, in their words, with the date. Below the table, every time a company actually changed those words. Over a thousand recorded versions of these documents come down to a few dozen real changes.
| Tool and plan | Trains on my prompts? | In their words | On record since |
|---|---|---|---|
| ChatGPT Free, Plus and Pro | Yes, by default. You can opt out. Data Controls in settings, or "do not train on my content" in OpenAI's privacy portal. Temporary Chats are not used for training. | When you use our services for individuals such as ChatGPT, Sora, or Operator, we may use your content to train our models.Privacy Policy | previous capture last seen |
| ChatGPT Business, Enterprise, Edu and API | No, unless you opt in. Opt-in only, for example through feedback mechanisms. | By default, we do not use your business data for training our models.Business Privacy Policy | first capture, may be older last seen |
| Claude.ai Free, Pro and Max | Yes, by default. You can opt out. Account settings. Conversations flagged for safety review and material you report as feedback are used even after opting out. | We may use your Inputs and Outputs to train and improve Anthropic AI models, unless you opt out through your account settings.Privacy Policy | last seen |
| Claude.ai Commercial (Team, Enterprise, API) | No. | Anthropic may not train models on Customer Content from Services.Commercial Terms | first capture, may be older last seen |
| Microsoft Copilot Consumer Copilot | Yes, in certain markets. You can opt out. Copilot settings. Which markets is not stated. | In certain markets, we use conversation data to train the generative AI models in Copilot, unless you choose to opt-out of such training.Privacy Policy · last good capture | previous capture last seen |
| GitHub Copilot Accounts under GitHub's Terms of Service | Yes, by default. You can opt out. Section J.3 of GitHub's Terms. Unless you opt out, GitHub's affiliates may also use your Inputs and Outputs. | If you provide your private repository content as Input to AI Features, we may use that Input to provide, develop, train, and improve the Service, including AI Features.Terms of Service | previous capture last seen |
| Perplexity API | No. | 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.Developer Terms | first capture, may be older last seen |
| Perplexity Enterprise (data processing agreement) | No. | For the avoidance of doubt, Personal Data will not be used for training of Perplexity's large language models.Data Processor Agreement | first capture, may be older last seen |
| Grok Consumer (grok.com, X and apps) | You choose in settings. The terms do not say which way it starts. A setting when logged in; the terms do not say which way it starts, but the apps ask you to choose at setup. Logged-out use grants xAI "full rights" to use your data for model training. | When logged into our Service, you can select whether or not you want us to use your User Content to improve our products and services and train our models.Terms of Service | previous capture last seen |
| Grok xAI API and enterprise | No. De-identified data derived from your use may still be used "for any lawful purpose". | For clarity, xAI shall not use any User Content to train any foundation models, large language models, or other artificial intelligence systems.Commercial Terms | previous capture last seen |
| Cursor All plans | No, unless you opt in. Training needs your explicit agreement. Preferences are managed in the app. | ANYSPHERE WILL NOT USE CONTENT TO TRAIN, OR ALLOW ANY THIRD PARTY TO TRAIN, ANY AI MODELS, UNLESS YOU'VE EXPLICITLY AGREED TO THE USE OF CONTENT FOR TRAINING.Terms of Service | previous capture last seen |
| Windsurf (Codeium) Free | Yes. No opt-out stated. None on the free tier. The opt-out is for paid tiers only. | Cognition may use Customer Data for model training purposes and to improve and enhance the Services.Terms of Service | previous capture last seen |
| Windsurf (Codeium) Paid (Pro, Teams) | Yes, by default. You can opt out. On Teams, only an administrator may opt out. | If you subscribe to a paid Service Tier, you may opt out of this use ("Opt-Out").Terms of Service | previous capture last seen |
| DeepSeek Consumer app and platform | Yes, by default. You can opt out. Listed among privacy rights: "the right to opt-out of using your Personal Data for training our models or optimizing our technologies." | To improve and develop the Services and to train and improve our technology, such as our machine learning models and algorithms.Privacy Policy | last seen |
| Jasper All plans | Yes. Opt-out only in some regions. EU and UK users only, by request to [email protected] . The licence covers "modifying, improving, and enhancing artificial intelligence models". | For users located in the European Union or United Kingdom, you may have the right to object to the use of your Customer Property for the purpose of improving artificial intelligence models.Terms of Service | previous capture last seen |
| Le Chat Free, Pro and Student | Yes, by default. You can opt out. Account settings. Feedback and moderated content are used regardless. | We do not use Your Data to train our artificial intelligence models except (a) when you use the Mistral AI Products under a free subscription, Le Chat Pro or Le Chat Student and you have not opted-out of trainingTerms of Service | first capture, may be older last seen |
| Le Chat Team and Enterprise | No, unless you opt in. Exceptions are feedback you give and content flagged by moderation. | We do not use Your Data to train our artificial intelligence models exceptTerms of Service | first capture, may be older last seen |
Latest changes #
Cognition now says it "may use Customer Data for model training purposes"; "paid Service Tier" customers "may opt out" (Teams administrator only). Change of position · Windsurf (Codeium), Terms of Service #
"Train our models" became "train and improve Anthropic AI models", and research on the societal impacts of AI became a stated purpose. Scope change · Claude.ai, Privacy Policy #
xAI's enterprise terms dropped automatic 30-day deletion of User Content and claimed ownership of de-identified data for any lawful purpose. Change of position · Grok, Commercial Terms #
Grok added a provision stating that it "shall not use any Google Apps content" connected via Google OAuth for "any of its internal AI or other training purposes." Scope change · Grok, Privacy Policy #
OpenAI now says, "if training is enabled" or "if you provide feedback," it uses content, and also uses "data from third-party partners" to train models. New disclosure · ChatGPT, Privacy Policy #
GitHub now says it may train on “Your Content” and Affiliates may use Inputs/Outputs “Unless you opt out,” replacing “if you allow in your settings.” Change of position · GitHub Copilot, Terms of Service
source & further reading
trainedon.me — original article