NeurIPS 2025: Detailed Policy on Penalties for Missing Reviews (Including Official AC Email Text)

root

Introduction

This report summarizes and shares the exact regulation and enforcement details for missing reviews at NeurIPS 2025. The information is based on the latest guidance distributed to Area Chairs (ACs) and reviewers, including the full transcription of the official NeurIPS 2025 AC Reminder email. This is to ensure the CS research community is fully informed of the responsibilities, penalties, and procedural changes related to peer review obligations at NeurIPS 2025.

Summary of NeurIPS 2025 Policy on Missing Reviews

Strict penalties will be imposed on reviewers who fail to submit required reviews or submit reviews of insufficient quality.
Reviews that are flagged as poor (e.g., too short, LLM-generated, or otherwise problematic) may result in desk rejection of the reviewer's own submissions.
Emergency reviewing protocols are in place, but abuse or neglect can also result in significant penalties.
At the meta-review stage, ACs will be asked to evaluate reviewer interactions and flag any grossly irresponsible behavior.

Exact Transcription of Official AC Email

Below is the exact transcription of the official NeurIPS 2025 Area Chair (AC) Reminder email, as requested:

[NeurIPS 2025 AC] Reminder: Responsible Reviewing (Final Reminder) and Emergency Reviewing

Dear XXXXX,

Thank you for serving as AC for NeurIPS 2025.

We are now less than 24h from the reviews due date -- the 2nd July 11.59pm AOE. We would like to take this opportunity and remind you to allocate sufficient time to aid the review process.

The conference key dates can be found here:
https://neurips.cc/Conferences/2025/Dates/

Please note we will be strictly imposing Responsible Reviewing this year with the active help of you, Area Chairs, and Senior Area Chairs.

The info about Responsible Reviewing initiative can be found below:
https://blog.neurips.cc/2025/05/02/responsible

Please note the following key points (see extra information re. Emergency Reviewers in Point 3):

We will be providing ACs with a FlagInsufficientReview button in OpenReview to flag poor quality, extremely short, aggressive, impolite, possibly LLM-generated, or other problematic reviews.

Reviewers recruited after 17th of June will be categorized as Emergency. Please do not use this functional on Emergency Reviewers unless an absolutely irresponsible review is received. Also, please be understanding if a reviewer was assigned late in the review process.

If a review is identified by this tool as insufficient, you are asked to prompt a reviewer to improve and resubmit the review by clicking the button, selecting "Request improvement of review", explaining the issue, and clicking "Submit". You can edit and submit this form as many times as needed.

Reviewers are obliged to try to improve reviews and communicate their efforts with you promptly.

Resolved reviews that were flagged require you to "Edit" the form, select radiobox "Review is satisfactory" and click on "Submit".

In addition to the button, later during the meta-review stage we will ask you to rate each reviewer based on your entire interaction with them and flag grossly irresponsible reviewers who, upon our review, may have all their own submissions desk-rejected from NeurIPS'25 as an ultimate penalty.

Reviewers are expected to maximize the quality of reviews subject to their expertise and confidence level. We ask them to work with you, communicate clearly and openly, and only use OpenReview if their confidence is not high, while doing their best to submit meaningful reviews. Lack of time or expertise are not valid reasons for not submitting or submitting low-quality reviews (there was sufficient time given already to communicate such issues to Area Chairs).

At this point we advise against unassigning any reviewers who cannot provide meaningful reviews due to genuine lack of expertise or time. We advise you (AC) to request them to do their best and indicate lower confidence in the review form while you should seek immediately an extra qualified Emergency Reviewer to be co-assigned. For other exceptional circumstances we provided a form for reviewers that requires submitting the description of circumstances & evidence for our review (closes on the 5th of July, 11:59pm AOE):
https://forms.gle/FJ1F3xkcS9VxuSWz5

Secondly, reviewers who fail to provide all reviews by the 2nd July 11:59pm AOE will be informed and their own co-authors will also be prompted about upcoming reviews when the rebuttal period begins.

To be removed from the blacklist, such reviewers have to communicate the timeline for the review with you (AC), and provide high quality reviews as soon as possible. If any review is missing by the 22nd July 11:59pm AOE, the reviewer and their co-authors will lose the access to reviews of their own submitted NeurIPS'25 papers.

If you need to add Emergency Reviewers at any point, we ask you to ensure the total number of reviews does not exceed 5 (otherwise it may overwhelm authors).

You are encouraged to choose Emergency Reviewer(s) from the list of existing Reviewers (preferably those with few assigned papers). You may also try to compose the standard reasons for inviting Emergency Reviewers (note that removing reviewers is discouraged):
https://docs.google.com/document/d/2PACX-1vRwX5vE88kJd1bqDCmfUnVlhJCEtI2iAc6nofiah9k_IK6fMtaMJ0tJ7GAxwU3BC3_oRm4XqefY/pub

You are also welcome to leverage your contacts who are currently NeurIPS reviewers or invite external reviewers (if their expertise is required) with the approval of your SAC. In those circumstances, please check first if they are willing to serve as Emergency Reviewer for any given paper. Please do not share exact details of the paper in case of COI. Sharing a more general area to make sure the Emergency Reviewer is interested is OK.

Emergency Reviewers should complete their work As Soon As Practically Possible and no later than 22nd July, 11:59pm AOE.

When SAC invites an external Emergency Reviewer, they will have to create an OpenReview profile if they do not have one. The system will then check for COI and only then assign the paper to them.

Please contact your Senior Area Chair if you would like help (e.g., any issues with reviewers) or suggestions with emergency reviewer assignments. Your Senior Area Chair contact appears on the Area Chair Console immediately below "Modify Reviewer Assignments" (look for "Your assigned Senior Area Chair").

The rest of this email contains excerpts from the previous message to ACs:

A reminder that your paper assignments can be accessed here:
https://openreview.net/group?id=NeurIPS.cc/2025/Conference/Area_Chairs

Last but not least, we have identified multiple instances of suspicious behaviors such as formation of co-bidding rings and are actively monitoring these cases. Any form of collusion in reviewer assignments or other aspects of the review process may result in immediate desk rejection of all papers associated with the individuals involved.

Thank you for helping NeurIPS 2025 and best wishes,

Key Points for the Community

All reviewers must submit meaningful, high-quality reviews on time. Failing to do so may result in desk rejection of your own submissions and loss of access to peer reviews.
Reviews identified as insufficient (poor, LLM-generated, impolite, or unreasonably short) may result in immediate sanctions, including being blacklisted.
Emergency Reviewers have a clear deadline (no later than July 22, 11:59pm AOE) and must also deliver high-quality reviews.
Any collusion or formation of review rings is strictly forbidden and will result in desk rejection of all associated papers.

References

NeurIPS 2025 Responsible Reviewing Initiative: https://blog.neurips.cc/2025/05/02/responsible
Area Chair resources and assignments: https://openreview.net/group?id=NeurIPS.cc/2025/Conference/Area_Chairs
Circumstances/Evidence Submission Form (closes July 5): https://forms.gle/FJ1F3xkcS9VxuSWz5
Emergency Reviewer Instructions: https://docs.google.com/document/d/2PACX-1vRwX5vE88kJd1bqDCmfUnVlhJCEtI2iAc6nofiah9k_IK6fMtaMJ0tJ7GAxwU3BC3_oRm4XqefY/pub

Please circulate this document within your research networks to ensure all potential NeurIPS reviewers and authors are fully aware of these strict new enforcement policies.

This was originally shared by 'George Lee'

CSPaper: peer review sidekick