Overleaf Terms Scorecard

Grade A- — Mostly protective (score 86/100)

Overleaf provides a standard service agreement that includes broad termination rights and forced arbitration, but explicitly pledges not to train AI models on user content without consent.

Standard terms. No data sale. Broad termination rights. Arbitration required.

Terms at a glance

Sells your data
NO
Forced arbitration
LIKELY
Trains AI on your content
NO
Collects biometric data
NO
Shares data with government
YES
Easy to cancel
YES
Tells you before changes
NO
Deletes data on request
YES

What the terms actually say

TERMINATION

Unless otherwise agreed, we reserve the right to suspend or end the Services at any time, with or without cause, and with or without notice.

Overleaf reserves the right to terminate your account at any time, for any reason, without providing you with prior notice.

LIMITATION OF LIABILITY

IN NO EVENT WILL OVERLEAF... BE LIABLE FOR... AN AMOUNT IN AGGREGATE THAT IS MORE THAN THE GREATER OF £0.01 OR THE AMOUNTS PAID BY YOU TO OVERLEAF FOR THE PAST THREE MONTHS

The company limits its total financial liability to you to a maximum of the amount you paid in the last three months, or one penny, whichever is greater.

OVERLEAF AI

We will not use your Inputs or Outputs to train or develop AI models except where this is part of the Services we provide to you.

Overleaf commits to not using your user-generated content for training their AI models without your explicit consent.

CHANGE TO THESE TERMS

We may revise these Terms from time to time... Other changes may be posted to our news or terms page, so please check those pages regularly.

The company does not guarantee email notification for all changes to the terms, placing the burden on the user to check for updates.

MISCELLANEOUS LEGAL TERMS

Both parties submit to the exclusive jurisdiction of the English courts

Disputes must be resolved in English courts, which may be a significant barrier for international users.

Last reviewed 2026-08-18 under rubric v3.5.

Other scorecards