# Privacy and data governance
URL: https://docs.clearly-ai.com/docs/privacy-and-data-governance
Description: Understand what data Clearly AI processes and how to request export or deletion.
Clearly AI processes the data your organization provides so it can run reviews, answer questions, and automate workflows.

## Data Clearly AI processes [#data-clearly-ai-processes]

Customer-provided data can include:

* Uploaded documents, diagrams, spreadsheets, PDFs, and text snippets.
* Code, tickets, pages, and files pulled through integrations.
* Knowledge Base content.
* Review Templates, review answers, citations, exports, risks, and workflow outputs.
* Business contact data used for authentication and access control, such as name, business email, team, and organization membership.

## AI data use [#ai-data-use]

Clearly AI does not use customer data to train or fine-tune model providers unless explicitly requested for customer-specific models.

Clearly AI uses Retrieval Augmented Generation so review and chat responses can use your sources and Knowledge Base context without training a model on that data.

For model provider controls, see [Model providers](https://docs.clearly-ai.com/docs/model-providers).

For broader security, infrastructure, compliance, and vendor governance controls, see [Security and privacy](https://docs.clearly-ai.com/docs/security-and-privacy).

## Export or deletion requests [#export-or-deletion-requests]

An authorized customer representative can request data export or deletion by emailing [security@clearly-ai.com](mailto:security@clearly-ai.com).

Include:

* Your organization name.
* The data to export or delete.
* Any relevant project, review, source, or user details.

When customer data is deleted, Clearly AI removes it from primary storage, removes related index references, de-indexes it from Knowledge Base retrieval, and records the deletion for compliance purposes.

Security, access, and authentication logs are retained for monitoring and compliance.

## Admin practices [#admin-practices]

* Upload only the sources needed for the review.
* Remove stale sources from projects and the Knowledge Base.
* Use source categories and trust levels to separate authoritative policy material from lower-confidence evidence.
* Use user-scoped credentials for personal access and organization-scoped credentials only for shared service accounts.

## Related [#related]

* [Security and privacy](https://docs.clearly-ai.com/docs/security-and-privacy)
* [Knowledge Base](https://docs.clearly-ai.com/docs/knowledge)
* [Integrations](https://docs.clearly-ai.com/docs/integrations)
* [Model providers](https://docs.clearly-ai.com/docs/model-providers)