AI tool configurations
This page is guidance, not part of your agreement. Clause 6.4 of the Data Licence Agreement requires you to use an AI tool configured so that your inputs are not used to train or improve any model. This page helps you work out whether yours is. The obligation, and the checking, remain yours.
Why this matters
When you use our data through an AI tool, your questions and our data go to that tool's provider.
Some AI plans learn from what you type into them. If yours does, our data can end up inside a model permanently, where neither you nor we can retrieve it and where it may surface for other people.
That is why the agreement asks you to use a plan that doesn't learn from inputs.
The one distinction that matters
It is not about which company you use. Across every major provider, the same line runs through the middle of their product range:
| Trains on your inputs? | |
|---|---|
| Consumer plans — free, personal, individual paid tiers | Often yes, by default, unless you turn it off |
| Business, team, enterprise and API plans | Generally no, by default, under commercial terms |
If you take one thing from this page: use a business or API plan, not a personal one. That single choice resolves the requirement for most people.
What providers currently publish
Checked 11 August 2026. Providers change these terms — treat this as a starting point, not an answer.
| Provider | Plan type | Position as published |
|---|---|---|
| Anthropic (Claude) | API, Claude for Work — Team and Enterprise | Inputs and outputs from commercial products are not used for training by default |
| Anthropic (Claude) | Free, Pro, Max | Governed by consumer terms, with a model-training setting the user selects. Check your privacy settings |
| OpenAI (ChatGPT) | Team, Enterprise, API | Business and API data is not used to train models by default, under commercial terms |
| OpenAI (ChatGPT) | Free, Plus, Pro | Governed by consumer terms, with a data-controls setting the user selects. Check your data controls |
| Other providers | Business, enterprise, API | Most publish a no-training commitment on commercial tiers. Confirm on your own account |
| Other providers | Free and personal paid | Frequently train by default, sometimes with an opt-out. Confirm on your own account |
Two exceptions that catch people out. Submitting feedback (a thumbs-up, a bug report, a survey) often sends your conversation down a different path with different retention, even on a business plan. And logging into a business tool with a personal account puts you on consumer terms regardless of what your organisation pays for.
How to check your own
- Find which plan your account is on — not which product, which plan
- Look for a setting named something like model training, improve the model, or data controls
- Confirm it is off, or that your plan's terms exclude training by default
- If you cannot tell, ask your provider's support in writing and keep the reply
What to tell us
At signup, and again at each renewal, tell us:
- the AI tool you use to access the Service
- the plan tier you are on
We don't ask for your account details, your prompts, or anything about your clients.
If your configuration changes
Tell us at support@datastandard.pro. If you move onto a plan that trains on inputs, stop using the Service through that tool until it is resolved.
The limits of this page
We have no visibility into your AI tool and no ability to verify your configuration. We publish this because it makes a contractual obligation easier to meet, not because we can stand behind any provider's terms.
This page changes as providers change their policies. It does not form part of the Data Licence Agreement, and nothing on it varies clause 6.4.
Last checked 11 August 2026 · support@datastandard.pro