ACL's New Policy: Submissions Must "Come with a Reviewer" — The Sustainable Reviewing Policy Explained

A New Solution to the Conference Submission Explosion

As large language model and NLP research continues to surge, top academic conferences are facing an unprecedented flood of submissions. The number of available reviewers has fallen far behind the exponential growth in submissions, and declining review quality along with reviewer burnout have become pain points for the entire community. Recently, ACL (the Association for Computational Linguistics) officially announced on X a proposal called the "Sustainable Reviewing Policy", aiming to address this structural imbalance at the institutional level.

The proposal sparked lively discussion in Reddit's machine learning community. Its core logic is straightforward: reviewing capacity should match submission volume, and every submission must "pay its own way." In other words, submitting a paper is no longer a zero-cost act — it requires a corresponding contribution of reviewing service.

Core Elements of the ACL Sustainable Reviewing Policy

The "Bring-Your-Own-Reviewer" Mechanism: Submissions and Reviewing Deeply Linked

The most striking aspect of the proposal is this: every submitted paper must provide one qualified service contributor (a reviewer or area chair). This means at least one member of the submitting team must be willing and qualified to take on reviewing responsibilities — effectively "paying" for the review resources their paper consumes.

If a submission cannot identify a qualified service contributor among its authors, it will be placed in a lottery pool, competing only for leftover reviewing capacity. This directly ties the ability to receive a review to a willingness to contribute, fundamentally reshaping the incentive structure around submissions.

Non-Author Nominations and the Endorsement Mechanism

Recognizing that not every team includes a qualified senior reviewer, the proposal also introduces a flexible alternative. If no author qualifies as a service contributor, teams may nominate a designated contributor who is not an author to fill this role. However, the nominee must formally "endorse" the work — a mechanism described as similar to arXiv's endorsement system.

Additionally, for researchers who have not yet reached reviewing eligibility — such as students new to the field — ACL plans to establish a mentorship system to help them grow into qualified reviewers over time. This design prevents the policy from becoming a pure seniority barrier, preserving a pathway for newcomers.

ACL Submission Quotas: Hard Limits Per Author

Beyond the reviewer-binding mechanism, the proposal also introduces per-author submission quotas:

• Total submission cap: 20 papers per cycle
• First-author submission cap: 5 papers per cycle (including co-first authors)

These quotas are intended to curb the practice of "shotgun" mass submissions by certain researchers. Many Reddit commenters noted that these numbers are actually quite generous — for the vast majority of researchers, submitting 5 first-author papers in a single cycle is already a remarkably high output, and hitting the 20-paper total ceiling is nearly impossible in normal practice. The quota primarily targets the rare accounts that systematically flood submissions.

Anti-Abuse and Penalty Mechanisms: Building an Academic Reputation System

To prevent the policy from being gamed, ACL has explicitly stated that anti-abuse measures will be enforced. Specifically:

• Accounts that systematically submit or endorse low-quality work will face penalties;
• Accounts that abuse the system in other ways may be penalized or even banned.

This signals that ACL is not only trying to guide good behavior through institutional design, but is also prepared to enforce consequences. The combination of endorsement accountability and penalty mechanisms essentially creates a reputation system: those who contribute quality reviews gain priority access to submissions, while those who abuse the system pay the price.

Community Reaction: Support and Controversy Side by Side

Supporters: Necessary Gatekeeping

In the Reddit discussion, many respondents expressed support for the proposal. The original poster stated plainly: "This seems quite reasonable to me — especially given that a large number of submissions come from teams where nobody is even qualified to review. It does have a bit of a gatekeeping flavor, but I think it's highly necessary."

This view reflects a widespread community concern: peer review is a shared public resource, and free public resources tend to be overexploited. When submitting is nearly costless but reviewing is something no one wants to do, the system is bound to collapse. ACL's proposal essentially introduces a property-rights and cost structure into this "commons," realigning those who contribute with those who benefit.

Potential Concerns and Challenges

Of course, this type of policy is not without risks. First, how will "qualified reviewer" status be defined? If the bar is set too high, it could exclude small teams with limited resources or researchers from less-resourced regions. Second, while the lottery mechanism preserves some fairness, it could also mean that high-quality work from teams without a reviewing contributor misses out on evaluation. Third, how the endorsement mechanism will prevent "favor endorsements" and collusive mutual endorsements in practice will require careful rule design.

That said, given that ACL has already built in flexible channels like the mentorship system and non-author nominations — and has clearly articulated anti-abuse penalties — these concerns are at least partially addressed.

An Experiment in Academic Community Sustainability

ACL's Sustainable Reviewing Policy can be seen as an institutional innovation experiment by a top academic conference grappling with exponential submission growth. Its core idea — "submissions must pay their own way" — transforms peer reviewing, long treated as a voluntary obligation, into a quantifiable, accountable form of contribution.

In an era when AI research has entered a period of "paper inflation," this kind of mechanism may well become a template for other top venues (such as NeurIPS, ICML, and CVPR) to adopt. Whether it can truly achieve "sustainability" depends on the fairness and flexibility of its implementation. But at the very least, ACL has taken a proactive step toward acknowledging the problem and pursuing reform.

ACL has stated that more detailed policy information will be released soon on its official website and social media channels.