Privacy Policy

The Civil Discourse Digital Toolkit is a free educational project of the University of Toronto, Institute of Medical Science. This page explains what happens to the text you enter and the limited information we keep.

What you type

The interactive tools send the text you type (prompts, drafts, and scenario responses) to third-party AI providers through the Vercel AI Gateway and OpenAI in order to generate a response or a score. We do not store the content of these requests on our own servers, and the public tools do not require an account.

Do not enter patient-identifiable information or protected health information (PHI). The examples here are teaching scenarios. Treat anything you type as if it will be processed by an external service.

AI processing

Generated text comes from large language models operated by our providers and is subject to their terms and privacy practices. Output can be inaccurate or incomplete and should not be relied on for clinical, legal, or other professional decisions.

Administrators

People who edit site content sign in with an email and password managed by Convex Auth. We store that account record and a session cookie solely to operate the admin dashboard. Admin access is limited to an allow-list and enforced on every request.

Cookies and analytics

The public site uses only the cookies needed for it to function. We do not sell personal information.

Your choices and contact

Because the public tools store no account or request history, there is little personal data to access or delete. For questions about this policy, contact the program through the University of Toronto Institute of Medical Science.