StealthGPT claims it "killed" AI detector Pangram with new humanization model Super

TuhinChakr · x · 2026-09-09

StealthGPT announced Super, a next-gen humanization model it claims can fool Pangram, "the world's most accurate AI detector," with a demo. Scholar Brendan Nyhan amplified it while asking how long until Pangram strikes back—escalating the AI-detection arms race.

Original post →

More from Fun

Fun channel →