Local work stays local by default

Browser-local decision data

Saved models, upgrade-radar settings, task sets, pasted answers, judgments, and draft reports are stored in browser storage on the device where you created them. They are not silently copied to an OpenGPT account. Clearing site data or using another browser can remove or hide this information.

Account identity and submissions

When you use an account-linked feature, OpenGPT can receive a site-specific user ID plus the name and email supplied by the authentication platform. OpenGPT cannot use that sign-in to read your chats or other private account content.

Public-data submissions are tied to the submitter so reviewers can audit abuse and evidence. Submitter identity is retained for at most 365 days. The account workspace provides export and anonymization controls; anonymization closes pending submissions while public evidence may remain without identity.

Analytics

The production site loads Google Analytics. OpenGPT sends a limited set of allowlisted product-event names and short identifiers or counts to understand whether public workflows work. The product event helper accepts only constrained keys and values.

• Do not send search text, prompts, answers, report contents, email addresses, file contents, credentials, or API keys as analytics events.
• Analytics is not loaded on localhost or unrelated hosts.
• Standard network and browser information may still be processed by the analytics provider according to its own service terms.

Public links and downloads

Files you choose in the report verifier are read locally and are not uploaded by that tool. Links to provider documentation, repositories, and benchmark publishers take you to third-party sites with their own privacy practices.

Questions and requests

For a privacy question or a request concerning account-linked submission identity, contact contact@opengpt.com. Use the in-product export and anonymization controls when available.