Pangram accused of scoring QualiaQuanta’s older papers 100% AI-generated

QualiaQuanta alleged that Pangram scored papers they wrote in 2016 or earlier as 100% AI-generated and warned against trusting AI to judge AI-written text. Jesse Singal later said a detector relying on Pangram rated a whole article by QualiaQuanta 100% human. Singal asserted that shorter snippets are a known problem for Pangram, but neither post establishes whether QualiaQuanta tested the papers in full.