Only you and anyone you explicitly invite. QED staff may access them for operational purposes (e.g., ensuring it was uploaded properly) or quality assurance (e.g., assessing our Critical Thinking AI performance). Our AI providers are contractually barred from training their own foundation models on your data.

QED will never publish, republish, or assert authorship over your unpublished work, nor interfere with any journal copyright you later choose. You retain full ownership of all manuscripts and data. You grant qed a revocable, non‑exclusive worldwide license solely to (i) generate feedback/reports, (ii) optionally train and evaluate QED's own models, and (iii) create anonymised, aggregated analytics.

No. QED is a private, pre‑submission tool. Reports stay inside your account unless you download or share them, so journal “originality” rules and copyright transfer agreements are unaffected (If a journal has a unique policy, we recommend checking its guidelines, just as you would with any lab notebook or pre‑print).

We strive to maintain industry standard data protection and security tools. We are extremely careful in how we use unpublished data in training. Your manuscript can indirectly affect our model parameters, but you are always in control of how your work gets published to the world.

Deleting a manuscript immediately ends qed’s licence to use it. All files are purged from live systems and scheduled for permanent erasure from encrypted backups within ≤ 30 days. Any model parameters already learned or anonymized stats remain because they cannot technically be rolled back, but they contain no identifiable text.

We log basic usage metrics (clicks, dwell‑time, ratings, etc.) to improve features, monitor security, and produce fully anonymised product‑analytics that may be shared or licensed. We never sell or share raw manuscripts or deleted content for marketing or any other purpose.

QED does not use submitted manuscripts to train it’s models to reproduce, rephrase, or refer to unpublished work, and we take strict measures to ensure that no unpublished content can be exposed or reconstructed in whole or in part unless the author explicitly chooses to share it through the qed platform. Submitted manuscripts may be used as input to improve QED’s ability to assess whether its feedback is helpful or likely to be acted upon.

Add the following sentence to the acknowledgements section:

Pre-submission review was conducted using qed Science (https://www.qedscience.com)

‍

No. qed assumes that the data and results are genuine. Its current purpose is not to investigate fraud but to analyze, validate, and interpret the scientific reasoning and evidence presented in the work. In the future, additional capabilities will allow qed to flag signs of manipulation.