2026Q3 Reports: ARR Editors in Chief
ARR Team
- Editors:
We recruited four new editors in late 2025 (Tanmoy Chakraborty, Parisa Kordjamshidi, Giovanni Da San Martino, and Asad Sayeed). This is done through advertisement of the role. Diversity of locations and coverage of all continents has been one of the major criteria as well as suitability and seniority of the candidates.
The current editors are Vincent Ng, Anna Rogers, Jing Jiang, Michael White, Margot Mieskes, Xiaodan Zhu, Sarvnaz Karimi, Xiaojun Wan, Giovanni Da San Martino, Tanmoy Chakraborty, Parisa Kordjamshidi, and Asad Sayeed. (12 in total)
- Tech Team:
In 2025, two of the ARR CTOs stepped down (Sudipta Kar and Jonathan K. Kummerfeld). We have recruited one new CTO (Yuntian Deng) to the Tech team who started in January 2026. One of the CTOs (Freda Shi) will step down in July 2026 due to personal reasons and training is underway for the new CTO.
- Communication & Support:
Shubhra Kanti Karmaker (Santu) leads the comms and support team and recently a number of members were added for a new term (https://aclrollingreview.org/organization).
- Paid positions within ARR:
ARR team is grateful for the four positions that have been created (two editorial assistants, workflow manager, a part-time workflow assistant from Openreview). These roles have enabled us to work through deadlines and manage the load. Noting one of the editorial assistants started in March 2026 cycle (Thomas). His support on handling administrative issues, keeping track of to-dos and timelines was (and continues to be) very valuable and helpful.
Submissions and Reviews
ARR has seen a large number of submissions each cycle. January 2026 saw over 10k submissions and May was over 17k+. This has put strain on the team handling the reviewing process. This problem is partly due to AI-generated papers. However, detection of these and desk rejection adds workload to the team. We are in communication with PEC and in continuous discussions to explore pathways to make this process sustainable. Some of the potential avenues are:
- Submission caps;
- Karma system (developed in a system led by Mausam/ex-ARR EiC) where some credit building based on experience would recognise good science citizenship/good reviewing behaviour as well as balancing the number of submissions.
- REVAS system: We are experimenting a new tool to assist the reviewers. The tool and how it is used is explained here: https://aclrollingreview.org/revas-may26
- Paper matching and assignment: As of January 2026, paper matching and assignments now taken over by the ARR tech team from OR.
- Hallucinated References: In January cycle, hallucinated references were detected post commitment with some of the accepted papers were then desk rejected when the references were identified using a third-party tool. Similarly, papers with hallucinations in the May cycle are being desk-rejected.
Supporting multiple conferences per cycle
- For the first time, AACL and EMNLP (two ACL conferences) shared the same cycle in May 2026. Some lessons from this setup will shape the future decisions on how to manage such settings.
- Decision has been made to support COLING and NAACL 2027 in the same cycle in October. Coordination are underway to plan for this early.