From the Editor-in-Chief

EIC Open Forum Posts

Updates and messages from the Editor-in-Chief of Operations Research, originally shared through the INFORMS Open Forum.

[September 24, 2026]

Statement Regarding LLM Use

Statement from the OPRE EiC (Amy R. Ward) Regarding LLM Use

The Operations Research journal represents a community of people. The journal exists to help foster our intellectual inquiry and to encourage us to engage intellectually with each other. The high standards of scholarship maintained by the journal motivate authors to think deeply about their work – both its technical content and its presentation to a broad audience. The review process requires the AreaEd, AE, and reviewers to spend their time and energy thinking carefully about any paper submitted to the journal.

I am happy when the use of LLM tools supports the intellectual atmosphere of the community of scholars engaging with the journal. However, I also find reasons to be concerned. I share some of the worries articulated in the Leiden Declaration, and I observe some similarities with issues happening in the software open source community, as articulated in this Inside Rust Blog post on August 5, 2026. I also wonder what it means for LLMs to speed up the process of discovery when the time required for human understanding stays on its own timescale, an idea that comes up in this Terrence Tao video.

What does the current LLM tool landscape mean for the journal review process? I use the following principles, developed by the OPRE AI AreaEd Subcommittee, to guide my answers to the problems that we are seeing, discussed below.

• Human judgement must be central to producing and evaluating research.
• Scholarship standards for the journal should not be lowered.
• Human accountability and responsibility are critical for authors and reviewers.
• Policing LLM usage is not possible.

Problem: Paper Submissions may not indicate effort and understanding.

I used to assume that a submitted paper represented a serious effort on the part of an author, and could not have been written without hard-earned understanding of a problem. The burden was on the review process to respect that effort. LLMs render such an assumption invalid, and that potentially affects the mindset with which a review team approaches a paper.

Problem: Reviewer capacity is strained.

The ability to write papers more quickly results in more papers being submitted. There is a high mental cost to reviewing. The AreaEd, AE, and reviewers must make decisions regarding whether the research questions are good, and whether the solution approach is appropriate, and whether the technical arguments are correct. These decisions must then be written and articulated in a way that can be conveyed to the author. The temptation to outsource some of this work to LLMs is present.

What to do?

There is a reputational incentive structure in place for authors to submit papers in good faith and for reviewers, AEs, and AreaEds to do their review work in good faith. That has always been there, and that has not changed. The fact that we are a (relatively) small community means that I will continue to rely on this incentive structure.

The main changes the journal is doing are:

1. Reducing the amount of author feedback provided when a paper is desk-rejected by the AreaEd or AE;
2. Placing more emphasis on clear writing.

The first change allows desk rejections to be done more quickly, which increases AreaEd and AE capacity. The second change means that strong technical results do not guarantee a paper will advance to the next round. The results must be written in a way that is oriented towards human readers. This (hopefully) eliminates overly long AI-generated proofs, that are long because human judgement regarding what parts need more vs less explanation is missing.

Another change the journal is considering include adding more AreaEds and AEs.

There is no current plan to integrate LLM tools into the regular review process as a requirement, although there has been some success using such tools to help with the data and code review process. Practically, INFORMS integrating an LLM tool into its platforms that respects author confidentiality has its own challenges. The philosophy is that each member of the review team must own its recommendations and the reasoning for those recommendations.

Parting Thought

I am very hopeful that our community will manage AI tools in a way that maximizes their potential benefit and minimizes their negative effects. We are learning. I will continue to monitor how their use is affecting the journal, and doing my best to respond appropriately.

Note: I wrote this statement, with input from other humans. I did not use any AI tools to write this statement. I would like to thank the OPRE AI AreaEd Subcommittee consisting of David Brown, Rene Caldentey, Dan Iancu, Rouba Ibrahim, Raman Randhawa (chair), and Cong Shi. I would also like to thank Gal Mendelson and Dennis Zhang for discussions that have informed my opinions.

[January 12, 2024]

Dear INFORMS Community,

I am grateful to John Birge for his excellent steering of the journal the past 6 years. His are large shoes to fill.

I invite you to read the updated editorial statements for the journal (mine and the Area Editor statements), and view each Area's Associate Editors. See /page/opre/editorial-statement.

The main change to the manuscript submissions is the double-anonymous submission procedure. Author names are not shared with reviewers. The style templates for the journal, updated to be double anonymous, can be found on the submissions guidelines page, /page/opre/submission-guidelines. On the same page, you will also notice that the journal is receptive to shorter submissions, and aims to have shorter review times for shorter submissions. Shorter submissions will be published in the same style as longer submissions (no "technical note" in the header). Intellectual coherence should dictate the submission length.

The manuscript submission process will ask you to suggest 3 Associate Editors from the relevant Area and 5 reviewers.

Thank you for your support of the Operations Research journal.

------------------------------

Amy Ward, incoming Editor-in-Chief for the Operations Research journal
Rothman Family Professor of Operations Management
The University of Chicago Booth School of Business

------------------------------