AI safety is one of the most consequential — and most contested — topics in technology today. Claims about deepfake detection failures, AI-driven job displacement, energy consumption of training runs, and responsible scaling commitments shape public policy and investment decisions. But many of these claims lack rigorous verification.

This page collects 18 safety-related claims assessed by the Claimer community. Each includes an evidence chain with support and challenge sources so you can evaluate the strength of the underlying evidence yourself. Of these claims, 3 have veracity scores below 60%, indicating significant dispute or insufficient evidence.