"If AI writing is fine, disclose it": lab policy on AI-assisted writing sparks debate
ErnestoSolo6 · x · 2026-10-22
Paul Novosad argues on X that if AI writing is legitimate in some context, you should simply tell the recipient—"use AI when it's the best way to communicate, but disclose." He says this is standard practice in his lab. A reply pushes back that people hide AI use because academics post detector screenshots (like Pangram) to delegitimize work without real proof.
Related event: Economist Novosad: Using AI to Write Is Fine, Hiding It Is the Problem(14 posts)→