Scholars refuse to review papers flagged 100% AI-generated by Pangram

On Sept 2, Jessica Hullman shared her template for declining review invitations when Pangram flags an abstract as 100% AI-generated, arguing that disentangling author intent from AI-generated claims is not worth reviewers' time.

Confirmed

• Hullman refuses to review papers whose abstracts Pangram flags as 100% AI-generated, using a standardized reply template
• Sandro Ambuehl has sent similar refusals but deliberately omits detection details, fearing authors will tweak AI text just enough to evade detection—the less detectors reveal, the easier AI text is caught
• Hullman disclosed that although journal policy generally bars rejecting papers solely on AI-detection results, editors treat such flags as signals for heightened scrutiny of paper claims, which often surfaces further problems
• The practice drew discussion among scholars, including an exchange between Andrew Gelman and Hullman

Why it matters

• With reviewer shortages, an influx of AI-generated content adds burden; outright refusal is becoming a routine tactic for some scholars
• Ambuehl's secrecy strategy reveals an emerging cat-and-mouse game between detectors and writers
• The 'flag → deeper manual check' editorial workflow means AI detection, while not an official rejection basis, still shapes paper outcomes—raising concern about false positives

2026-09-02 ~ 2026-09-02 · 5 related posts

• [source] Professor declines review request citing 100% AI-generated abstract flag — JessicaHullman · 2026-09-02
• [source] Researcher declines AI-written submissions but hides detector details to prevent evasion — SandroAmbuehl · 2026-09-02
• [source] Editors treat AI writing flags as a signal for stricter scrutiny — JessicaHullman · 2026-09-02
• Reviewer now rejects papers flagged 100% AI-generated by Pangram detector — dggoldst · 2026-09-02

1 near-duplicate retellings: JessicaHullman