Acceptable use policy
What Dr. AI is for, what it is not for, the attestation every reviewer signs, and the controls that back it up. This is the shortest of our policies and the one worth reading.
Last updated 1 Aug 2026
1.Why this policy exists
Similarity and AI-detection tools sit at the centre of academic integrity processes. A service that connects people to those tools can support integrity or undermine it, depending entirely on the rules it enforces. This policy is where we draw that line, and it is enforced in the product, not just written down here.
It forms part of our terms and conditions. Every reviewer accepts it explicitly before approval, and clients accept it on registration.
2.Permitted use
- Checking work you wrote, or work you are otherwise authorised to submit, before you submit it.
- Understanding a report you have already received from an institution or publisher.
- Screening a manuscript for similarity before submitting it to a journal.
- Editorial, supervisory or institutional review of documents you are responsible for, with the author's knowledge.
- Teaching, training and research about how these tools behave, including their false positives.
3.Prohibited use
- Uploading a document you do not own and are not authorised to share.
- Asking for changes whose purpose is to disguise a document's origin, defeat a detection tool, or evade an academic-integrity process.
- Asking a reviewer to write, ghost-write, or substantially rewrite work that will be submitted as your own.
- Submitting another person's work in order to obtain a report about them without their knowledge.
- Attempting to identify, contact or pay the other party outside the platform.
- Uploading malware, or content that is unlawful in the jurisdiction of either party.
4.The reviewer attestation
Before any licensed-report service is dispatched to a reviewer, they must accept the following, and an administrator must separately verify their evidence of access:
- I hold authorised access to the software I am offering to use, in my own right, and I will use it only within the terms of that access.
- I will not share, sell or lend my credentials, and I will not use an institution's licence for work it does not permit.
- I will run checks in a mode that does not add a client's document to any similarity repository.
- I will not accept an order whose purpose is to disguise a document's origin or to defeat a detection process, and I will report any such request.
- I will not attempt to identify or contact a client outside the platform, and I will not retain client documents after an order closes.
5.How the platform enforces this
- Services flagged as requiring licensed access can only be dispatched to reviewers whose capability for that specific service has been verified by an administrator.
- A reviewer with no accepted acceptable-use attestation is excluded from the candidate pool entirely — the exclusion is in the dispatch query, not in a reminder email.
- Workspace messages have contact details and off-platform payment requests removed before they are stored, and attempts are flagged for review.
- Author metadata removal exists so that a client's identity is not visible to the reviewer. It deliberately does not touch document content: institution names, course codes, supervisor names and title pages are left exactly as submitted.
- Every action that touches money, files or access is written to an append-only audit log.
6.A note on metadata removal
We remove author identity fields — document author, last-modified-by, company, custom properties, and tracked-change and comment authorship — from the copy a reviewer receives. That exists to keep the workspace anonymous in both directions.
We do not, and will not, strip identifiers from document content in order to make a document harder to trace. If you need blinded content for a legitimate reason, such as preparing a manuscript for double-blind peer review, do that yourself before uploading, and tell your reviewer that is what you have done.
7.Reporting a breach
Use the contact form and choose "Report misuse of the platform". Reviewers can also decline any order and give a reason; declining costs a reviewer nothing, and a decline citing this policy is escalated to our team rather than simply moving to the next reviewer.
8.Enforcement
- First-instance breaches by a client: the order is cancelled, the payment is returned, and a warning is recorded.
- Serious or repeated breaches: the account is closed and any balance is returned through the original payment route where possible.
- Reviewer breaches of the licensed-access attestation: immediate removal from the reviewer pool and forfeiture of the affected order's payout, pending review.
- Where a breach involves a third party's rights, we will respond to a properly evidenced request from that party.
If you are unsure whether your intended use is permitted, ask before you order. We would much rather answer the question than close the account.