>How do I tell my colleagues to stop contributing unverified AI output without creating tension between us?

Pattern rather than person? General team reviews or the like. As long as it's not tech leadership pressing for it..