Study confirms AI detectors systematically flag autistic writers' text as AI-generated

A George Mason University arXiv study using 60,000 Reddit posts confirms that AI detectors, including Turnitin, systematically misclassify writing by autistic and other neurodivergent authors as AI-generated due to low perplexity, alarming affected writers.

2026-10-19 ~ 2026-10-19 · 2 related posts

• Study: AI detectors disproportionately flag autistic writing as AI-generated — sanjanasinghx · 2026-10-19
• AI detectors flag neurodivergent writing as AI-generated, causing real harm — sanjanasinghx · 2026-10-19