Frequently asked questions (FAQ)

Can the AI be wrong?

Yes. AI answers can be incomplete or inaccurate. It works from the context you provide, so unclear questions or missing documents can lead to weak results. Always review important outputs and verify critical details.

Do my data train the models?

No. Your content is not used to train public AI models without your control. The assistant only uses the text and context you share in your conversation to generate a response.

Why is an answer incomplete?

This usually happens when the context is too large, too small, or missing the right sources. Try narrowing the question, adding the key documents, or splitting a request into smaller steps.

Technical limitations

Here are a few current limits to keep in mind:

• The assistant cannot read data outside the context you selected for that conversation
• It cannot access information from other workspaces in the same request
• Large files or long chats may require focusing on specific excerpts