Security & Data Retention
Last updated: April 29, 2026
No AI Training on Specs
QA Assist does not train AI models on your feature specs, Jira issue text, generated test cases, or reviewed plans. The app uses the Anthropic API for generation, and Anthropic states that commercial API inputs and outputs are not used for model training by default.
What Is Stored
- Signed-in users: saved generation history, projects, reviewed test cases, and review statuses.
- Guests: generated results are returned to the browser but are not saved to user history.
- Jira import credentials: used for the import request only and not stored by QA Assist.
Third-Party Processing
Feature descriptions are sent to Anthropic's API to generate test cases. Jira issue content is fetched from Atlassian only when a signed-in user provides a Jira URL and credentials for import.
Retention Notes
Saved plans are retained to provide history, project organization, and exports. Anthropic states that standard API inputs and outputs are deleted from its backend within 30 days unless an exception applies.
Recommended Customer Practice
Avoid pasting production secrets, passwords, private keys, or regulated personal data into feature descriptions or Jira imports. Specs often contain sensitive product context, so teams should follow internal data handling policies when using the app.