Privacy Policy

Most people arrive here with one of four questions, usually about an unpublished manuscript they are about to upload. The answers are below, and each one is expanded on further down the page.

CiteTrue is operated by DocumentMind LLC, registered in Cheyenne, Wyoming, United States. For the purposes of the GDPR we are the controller of your account data, and — when you submit work that is not your own — a processor acting on your instructions, under the DPA at /dpa.

CiteTrue checks whether the citations in a piece of academic writing actually exist and actually say what the text claims. This policy covers the website, the dashboard, the browser extension, the macOS app, the API and the MCP server — everything that runs under citetrue.com.

It does not cover the scholarly databases and search engines we query on your behalf; those services have their own policies, and what we send them is described in section 04.

Account data: your email address, authentication records, the profile details you choose to fill in (occupation, discipline, country), your plan and credit balance, and — if you pay — a customer reference held by Stripe. We never see or store full card numbers.

What you submit: the documents you upload and the references, DOIs, titles and URLs you paste in. Uploaded files are stored so that a finished verification stays reproducible when you open it again months later.

Results: the verification outcome for each reference, the sources that were matched, and the reasoning shown in the report.

Usage data: which features you used, credits consumed, API key usage, and ordinary server logs (IP address, user agent, timestamps) kept for abuse prevention and debugging.

This is the part worth reading closely, because it is the only part where your text leaves our servers.

When you submit a document, we extract its reference list. Each reference is then checked against public scholarly sources — Crossref and OpenAlex for bibliographic records, and web and search results when a reference has no DOI. What we send them is bibliographic: titles, author names, DOIs, journal names, years, and sometimes the reference string as written. We do not send your identity, your account, or the surrounding text of your manuscript.

To judge whether a matched record really is the work cited, and whether a claim is supported, we send the reference and the relevant passage to a commercial language-model provider. This is the only case where prose from your document is transmitted, and it is limited to the passage being checked — never the whole manuscript in one piece.

Uploaded files themselves stay in our storage. They are not handed to the model providers or to the scholarly databases.

We do not train models on your documents, your references or your results. We do not sell your data, we do not build advertising profiles, and we do not add your work to any plagiarism-detection corpus.

The model providers we call are OpenAI, Anthropic, Google and OpenRouter, over their commercial APIs. We may improve our own prompts and matching logic by looking at aggregate, de-identified statistics — how often a source type fails to match, for instance — never by reading your manuscript.

Account data, uploaded documents and verification results are processed and stored in the United States. If you are in the EEA, the UK or Switzerland, that is an international transfer; we rely on the European Commission’s Standard Contractual Clauses with the providers listed below.

We use the following providers to run the service. Each processes data only as needed to do its job. We will update this list before adding anyone new.

We keep what is needed to show you your own history, and nothing is kept because it might be useful later. You can delete an individual verification at any time from /dashboard, and you can delete your entire account from the same place — that removes your profile, verification history, credits and API keys.

We set one kind of cookie: the session cookie that keeps you signed in, plus a small preference cookie remembering your language. Both are strictly necessary for the site to work and are not used for tracking. There are no advertising or profiling cookies anywhere on this site.

We measure traffic with a privacy-friendly analytics tool that sets no cookies, stores no personal data and does not follow visitors across sites. It runs on our public pages only — never on the contents of your verifications.

Data is encrypted in transit and at rest. Access to production systems is restricted to the people who operate the service and is used for operating it, not for reading your work. Verification data is scoped to your account, and API keys are scoped to the account that created them.

No system is perfect. If we ever become aware of a breach affecting your data, we will tell you promptly and describe what happened rather than what we wish had happened.

Depending on where you live, you may have the right to access, correct, export or delete your personal data, and to object to certain processing. Most of this is self-service: your history and results are visible and exportable from /dashboard, and deletion is immediate.

For anything the dashboard does not cover — including a copy of everything we hold about you — write to [email protected] and we will respond within thirty days.

You decide what to upload. If a manuscript contains other people’s personal data, confidential material, or work you do not have the right to share, that judgement is yours to make before submitting it, not ours to make afterwards.

CiteTrue is built for researchers, students and professionals. We do not knowingly collect personal information from anyone under 13. If you believe a child has created an account, write to us and we will remove it.

If this policy changes materially, we will update the date at the top and email account holders before the change takes effect. Minor clarifications may be made without notice, but the sub-processor list is never quietly extended.

Questions about this policy, about what we hold, or about a specific verification: [email protected]. A real person reads that inbox.