HCI papers increasingly use LLM judges while obfuscating it, researcher warns

IanArawjo · x · 2026-09-03

Researcher Ian Arawjo reports finding more and more papers at HCI venues that use LLM judges, with a worrying trend: the writing obfuscates that LLM judges were used at all — sometimes it's hard to even tell what "IRR" (inter-rater reliability) refers to. The observation raises academic transparency and reviewing-norm concerns.

Related event: HCI Papers Quietly Use LLM Judges, Obscuring the Fact(2 posts)→

Original post →

More from Research

Research channel →