[D] SIGIR 2026 Reviews are (likely) done. Why the delay in releasing scores?

A growing debate within the AI research community centers on the delayed release of paper review scores for SIGIR 2026, a premier conference on information retrieval. With the official review deadline now passed, authors report that acceptance scores remain unpublished, creating uncertainty and hindering productivity. The viral discussion, led by researchers on platforms like Reddit, argues that the traditional practice of withholding scores for 'minor adjustments' by Area Chairs (ACs) is outdated. In today's accelerated research environment, where submission cycles for major venues like NeurIPS, ICLR, and ACL overlap tightly, each day of delay forces authors to wait in limbo instead of revising work or targeting alternative conferences.

The core proposal gaining traction is a shift toward 'rolling' or immediate score releases once all reviews are submitted. Proponents contend this transparency would not compromise the integrity of the peer-review process, as final acceptance decisions would still follow committee discussions. Instead, it would provide authors with critical early feedback on their paper's likely trajectory—be it acceptance, rejection, or borderline status—allowing for immediate strategic planning. This is particularly crucial for early-career researchers and students whose publication timelines directly impact graduation and job prospects.

The implications extend beyond SIGIR to the broader AI conference circuit. The current system often creates a cascade of delays, where authors waiting on one conference's results miss submission windows for others, compressing revision periods and increasing stress. Adopting a more transparent, rolling-release model could significantly improve research efficiency and mental well-being in the field. While some may argue it could lead to premature assumptions, the community sentiment suggests the benefits of empowering authors with information outweigh the risks, marking a potential inflection point for how AI conferences manage their most valuable asset: researcher time and momentum.