Does this vendor train AI models on your data? Per-product, per-tier, quoted from the current policy
For each software vendor and each plan tier, whether customer content is used to train the vendor's or a subprocessor's AI models, quoted verbatim from the vendor's own current terms, DPA or trust page, with the date the page was read. Answers 'does Slack train its AI on my messages', 'can I paste client code into Cursor', 'which of Grammarly, Notion, Figma and Atlassian train on customer content by default', 'is the enterprise tier different from the free tier', and 'how do I turn it off'. The answers genuinely differ and they differ by tier: Microsoft states flatly that 'Prompts, responses, and data accessed through Microsoft Graph aren't used to train foundation LLMs'; Notion states 'By default, Notion and its AI Subprocessors do not use Customer Data to train any models'; Figma's AI terms say 'When “Content Training” is toggled on within Customer's administrative user settings, Figma may use Customer Content to maintain, improve...'; and Grammarly's privacy policy says outright 'We also use information we collect to train our AI models' with an account-level control. Same question, four different answers, no page anywhere puts them side by side with the words that create the difference.
Records14
Sources12
Verified
Review by
LicenceEach record quotes a short operative clause from a publicly posted vendor policy and links to it. Facts about what a policy says are not copyrightable and the quotation is short, attributed and for the purpose of comparison; no policy is reproduced in whole. Vendor policies carry no reuse licence, which is the web's normal state.
The data
| Vendor | Product or service | Trains on your content by default? | What the policy says | Document | How it is controlled | Document date, as stated | Plan or tier |
|---|---|---|---|---|---|---|---|
| Slack | AI in Slack | opt-in | Slack will not use Customer Data to train generative AI models unless Customer provides affirmative opt-in consent. | Privacy Principles: Search, Learning and Artificial Intelligence | |||
| OpenAI | ChatGPT Business, ChatGPT Enterprise, ChatGPT for Healthcare, ChatGPT Edu, ChatGPT for Teachers and our API Platform | no | By default, data from ChatGPT Business, ChatGPT Enterprise, ChatGPT for Healthcare, ChatGPT Edu, ChatGPT for Teachers, and the API Platform (after March 1, 2023) isn’t used for training our models, unless you have explicitly opted in to share your data with us to improve the services. | Enterprise privacy at OpenAI | opt-in feedback mechanisms | Updated: January 8, 2026 | |
| Anthropic | Claude Free, Pro, Max | opt-in | We will use your chats and coding sessions (including to improve our models) if: You choose to allow us to use your chats and coding sessions to improve Claude | Is my data used for model training? | even if you have enabled Model Improvement in your Privacy Settings | March 16, 2026 | |
| Figma | Figma AI | opt-in | When “Content Training” is toggled on within Customer’s administrative user settings, Figma may use Customer Content to maintain, improve, and enhance Figma’s products and services by training machine learning and artificial intelligence algorithms and models. | Figma AI Terms | Content Training toggle | June 24, 2026 | |
| Gemini Apps | unclear | Google uses your activity to provide, develop, and improve its services (including training generative AI models), as well as to protect Google, its users, and the public with the help of human reviewers. | Gemini Apps Privacy Hub | If the Keep Activity setting is on: Your chats and what you share with Gemini (like files, videos, screens, and photos) will be saved in your Activity. | August 10, 2026 | ||
| Superhuman | Grammarly | yes | We also use information we collect to train our AI models. You can decide whether Superhuman can use your user content to train our AI models by adjusting the available training control(s) in your account settings. | Privacy Policy | the available training control(s) in your account settings | Effective as of July 6, 2026 | |
| Microsoft | Microsoft 365 Copilot | no | Prompts, responses, and data accessed through Microsoft Graph aren't used to train foundation LLMs, including those used by Microsoft 365 Copilot. | Data, Privacy, and Security for Microsoft 365 Copilot | 2026-07-09 | ||
| Notion | Notion AI | no | By default, Notion and its AI Subprocessors do not use Customer Data to train any models. | Notion AI security & privacy practices | |||
| Slack | predictive models for features such as emoji and channel recommendations | yes | We do not develop generative AI models using Customer Data. To develop predictive models for features such as emoji and channel recommendations, our systems analyze Customer Data (e.g. messages, content, and files) submitted to Slack as well as Other Information | Privacy Principles: Search, Learning and Artificial Intelligence | If you want to exclude your Customer Data from helping train Slack global models, you can opt out. | ||
| Atlassian | Rovo | no | Atlassian does not share customer metadata or in-app data with our third-party-hosted LLM providers for them to use to train or improve their services. | AI Trust | This use of contributed data is subject to data contribution settings, and we apply robust safeguards, including de-identifying and aggregating all contributed metadata before use. | ||
| OpenAI | ChatGPT Business, ChatGPT Enterprise, and our API Platform | no | By default, we do not train on any inputs or outputs from our products for business users, including ChatGPT Business, ChatGPT Enterprise, and the API. | How your data is used to improve model performance | We offer API customers a way to opt-in to share data with us, such as by providing feedback in the Playground | Updated: 12 days ago | Services for businesses |
| OpenAI | ChatGPT and Codex | yes | When you use our services for individuals such as ChatGPT and Codex, we may use your content to train our models. | How your data is used to improve model performance | You can opt out of training through our privacy portal by clicking on “do not train on my content.” | Updated: 12 days ago | Services for individuals |
| Anthropic | the Services | no | Anthropic may not train models on Customer Content from Services. | Commercial Terms of Service | Effective June 17, 2025 | ||
| Zoom | the Services | no | Zoom does not use any of your audio, video, chat, screen sharing, attachments or other communications-like Customer Content (such as poll results, whiteboard and reactions) to train Zoom or third-party artificial intelligence models. | Zoom Terms Of Service | Effective Date: August 11, 2023 |
Where this came from
Every record above links the page it was taken from and quotes the sentence that states it. These are the 12 sources this dataset was assembled from.
- slack.comhttps://slack.com/trust/data-management/privacy-principles
- openai.comhttps://openai.com/enterprise-privacy/
- privacy.anthropic.comhttps://privacy.anthropic.com/en/articles/10023580-is-my-data-used-for-model-training
- figma.comhttps://www.figma.com/legal/ai-terms/
- support.google.comhttps://support.google.com/gemini/answer/13594961
- grammarly.comhttps://www.grammarly.com/privacy-policy
- learn.microsoft.comhttps://learn.microsoft.com/en-us/copilot/microsoft-365/microsoft-365-copilot-privacy
- notion.comhttps://www.notion.com/help/notion-ai-security-practices
- atlassian.comhttps://www.atlassian.com/trust/atlassian-intelligence
- help.openai.comhttps://help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance
- anthropic.comhttps://www.anthropic.com/legal/commercial-terms
- zoom.comhttps://www.zoom.com/en/trust/terms/
Machine-readable
- data.jsonThe whole dataset — every record with its source URL and source quote.
- Open Knowledge Format bundleOne JSON object per line — every record's frontmatter and quoted span exactly as it is held here, in one fetch.
- How this is made and checkedWhat "verified against source" does and does not mean.