When the answer changed
Every time a company changed what its terms say about training on your prompts, newest first. Reworded sentences, link updates and archive glitches are left out on purpose.
- 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
- 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
- DeepSeek now states that "certain entities within our corporate group" process your provided, automatically collected and third-party data for "foundation model training and optimization"; the opt-out remains. Change of position · DeepSeek, Privacy Policy
- OpenAI now says “By default, data from ... ChatGPT for Healthcare ... isn't used for training ... unless you have explicitly opted in.” Scope change · ChatGPT, Business Privacy Policy
- DeepSeek still may train on Personal Data, but now grants “the right to opt-out of using your Personal Data for training our models or optimizing our technologies.” Change of position · DeepSeek, Privacy Policy
- Claude.ai (Anthropic) added use of Materials “to develop other products and services, including training our models,” “unless you opt out of training through your account settings.” Scope change · Claude.ai, Terms of Service
- OpenAI now says “By default, we do not train… on any business data accessed from connectors,” replacing “We do not train…” Scope change · ChatGPT, Business Privacy Policy
- Anthropic still says it may use “Inputs and Outputs” to train “unless you opt out through your account settings”; it adds “Legitimate interests” language. New disclosure · Claude.ai, Privacy Policy
- Microsoft now states that "in certain markets, we use conversation data to train the generative AI models in Copilot, unless you choose to opt-out of such training." Scope change · Microsoft Copilot, Privacy Policy
- Jasper still may use Customer Property for “modifying, improving, and enhancing artificial intelligence models,” but EU/UK users “may have the right to object.” Change of position · Jasper, Terms of Service
- Anthropic added that it may obtain personal data from “third party sources in order to train our models,” including feedback, flagged, or opted-in inputs and outputs. New disclosure · Claude.ai, Privacy Policy
- Anthropic's Claude.ai now uses Materials for "training our models, unless you opt out of training through your account settings," with feedback and safety-flagged content trained on regardless. Change of position · Claude.ai, Terms of Service
- Claude.ai (Anthropic) now says it “may use your Inputs and Outputs to train our models…unless you opt out through your account settings,” with safety-review/reported-materials exceptions. Change of position · Claude.ai, Privacy Policy
- Grok (xAI) now says logged-in users “can select whether or not” User Content trains models; logged-out use “where permitted” grants “full rights” for training. Change of position · Grok, Terms of Service
- OpenAI still says it "may use your content to train our models" for consumer services like ChatGPT, Sora and Operator "unless you opt out," now with added detail on opt-out and Temporary Chat. New disclosure · ChatGPT, Privacy Policy
- Microsoft still says it "may use your data to develop and train our AI models," but now states "Xbox's AI-enhanced features do not use children's data for model training." Scope change · Microsoft Copilot, Privacy Policy
- Cursor (Anysphere) says “We do not use Inputs or Suggestions to train our models unless” flagged for security review, explicitly reported, or “explicitly agreed.” New disclosure · Cursor, Privacy Policy
- DeepSeek now says it may use “publicly available information via the Internet sources” “to train our models,” while still training on interactions/usage. Scope change · DeepSeek, Privacy Policy
- GitHub Copilot says Free users’ data “may be used for AI Model training where permitted and if you allow in your settings.” New disclosure · GitHub Copilot, Terms of Service
- OpenAI now says it uses Content to train models powering “our Services,” not just “ChatGPT,” with an opt-out for Content. Scope change · ChatGPT, Privacy Policy