Editor declines peer review after Pangram flags abstract as 100% AI-generated

JessicaHullman · x · 2026-09-02

Columbia professor Jessica Hullman shared her new go-to review-decline template: she turns down review requests when the abstract is flagged 100% AI-generated by Pangram, since deciphering which claims came from the author versus AI "is not a good use of my time." She noted that at Science family journals, editors are restricted by policy from rejecting purely on AI detection, so they can't simply cite Pangram flags to authors. Economist Sandro Ambuehl replied he has sent similar messages but deliberately avoids naming detection details, worried authors could minimally modify AI text to evade detectors.

Related event: Professor's Rejection of Paper Flagged 100% AI-Generated Sparks Academic Debate(4 posts)→

Original post →

More from Safety

Safety channel →