Work in progress!

⚠️ This policy is new, and ARR is still working out the implementation details. We also receive a lot of questions. We will be posting any relevant clarifications and FAQ here and in the qualifications page (https://aclrollingreview.org/qualifications).

In light of challenges with unsustainable rate of submissions at EMNLP’26, ACL Peer Review Standing Committee has developed a proposal for a sustainable reviewing policy. It has now been approved by ACL exec team, and it will apply for submissions to ACL Rolling Review (ARR) starting from October 2026.

TLDR: ARR faces a reviewing crisis: submissions are growing far faster than reviewer capacity. 38% of the ~17K May 2026 submissions came from authors with no available publication record, and there are cases of outright spam, for which we so far had no established protocols.

The proposal: cap reviewed submissions to available reviewer capacity. Each submission must “pay” for itself by providing a qualified service contributor (reviewer or chair). Submissions without service capacity go into a lottery for whatever spare capacity remains. Service from qualified contributors counts towards the venue capacity, and a mentorship system will be built up for those not qualified yet. If a qualified service contributor is not available among the authors, non-author designated contributors can be nominated, but must vouch for the work (arXiv-endorsement style). Measures for system abuse will be implemented. For example, accounts that systematically submit or endorse low-quality work, and accounts that misuse the system in other ways, will be penalized or even banned. Additionally, per-author quotas are introduced, where authors are capped at 20 total submissions and 5 first-author (including shared first-author) submissions per cycle.

Offsetting measures: to keep things as inclusive as possible, reviewer qualification criteria are being broadened (already implemented in Aug 2026), with further proposals for a manual application path for those with equivalent experience, a mentor-matching program for newcomers, and a certification path to let experienced secondary reviewers become fully qualified.

Please see the full policy proposal for more details.

The new reviewer qualifications, capping of reviewing to available capacity, AND requirements for all authors to have completed OpenReview profiles including ORCiD come into effect immediately, and will apply to submissions to NAACL and COLING 2026 (ARR submission deadline - October 12).

We advise the following steps for those who are planning to make a submission for NAACL/COLING:

Make sure that ALL co-authors have OpenReview profiles including ORCiD - and, if available, DBLP/ACL Anthology links. It was previously possible for some co-authors to only be registered by email, but this will no longer be possible. Since it may take a while to coordinate with all authors, we recommend everyone to do this as soon as possible. The OpenReview profiles can be registered and updated before the official submission form opens.
Check the new reviewer qualifications, and decide who will serve as the designated service contributor. One qualified individual can serve on maximum two submissions (per ARR cycle, regardless of the commitment venue). Further make sure that this person has a full OpenReview profile with all publications, is qualified to review, and will not be named on more than two submissions. Any submissions with over-committed service contributors will be subset to max capacity randomly. The submissions that do not provide service capacity will enter a lottery, and may be assigned to reviewers or desk rejected based on available capacity. The same applies to the cases where the named service contributor fails to submit the registration form 48h after submission deadline at the latest.
The designated service contributor will need to vouch that the submission is at the expected quality level: ready for consideration for acceptance at a top-tier conference.
Those who are not yet qualified in terms of past publications, but have a stellar record of reviewing or secondary reviewing at ARR, will be able apply for manual verification. Early applications are highly encouraged, close-to-deadline requests for manual verification may not get processed on time.
Check the new service obligations: if a designated service contributor has any emergencies, the responsibility to notify about that and to provide a suitable and willing replacement is now on them. Otherwise any submissions they are serving for, and any they co-author, will be desk rejected.
Make sure that you are not planning more than 20 submissions, and not more than 5 as a first or joing-first author.

See this page for the most current ARR service qualifications criteria and instructions for how to check the qualifications.

ACL Peer Review Standing Committee https://www.aclweb.org/portal/content/acl-peer-review-standing-committee

Frequently Asked Questions

Could you include <venue X> in the qualifications checker?

We have already significantly updated the venue list from the policy. This page contains the most current list and instructions for how to apply for more venues to be included.

How many papers can a designated contributor support?

As described in the proposal, the maximum number of submissions per one contributor is 2. This entails max 10 reviews per person, which is already a relatively high service load. However, it is possible for non-authors to be nominated. The authors who need to submit more can leverage their networks to bring in more service contributors, decentralizing the recruitment process.

What if my designated service contributor is unavailable for filling the form 48h after deadline?

The contributor form becomes available in OpenReview console of the contributor, once a submission naming that contributor is made. This happens even if the submission is not yet final. The submission site normally opens 2 weeks before the deadline. So, you can make a submission with partial details, and have your designated contributor fill in the form at any time before the form deadline (48h after submission deadline).

Can’t you allow one ‘free’ submission for the new members of the community?

Unfortunately, the data suggests that we cannot. In 38% submissions in EMNLP 2026 that did not contribute service, most authors were submitting only one paper. But this already created a huge overload on others in the service pool, and the community response was very critical. However, lottery does allow us to review as many unsupported submissions as we can do sustainably. If the community steps forward with more volunteer service to support the new members, we will automatically be able to accept more papers for review.

Does this apply to demo or industry tracks?

No, this is only for the ‘main’ ARR review cycles. Demo and industry tracks are run separately, and may choose their own approaches.

Do the service contributors have to be in the same track as the submission?

No. E.g. it is possible to have a designated contributor qualified for IR track, even if the submission for which they serve is in another track.

Do resubmissions still need to bring in a service contributor?

Yes. Resubmissions still have to be considered by as many people as first-time submissions, and so they still need to bring in a service contributor, or be subject to lottery.