how far is too far when it comes to ai

Council verdict: We agree that AI goes “too far” when it (1) makes final, unreviewable decisions in dignity‑critical domains, (2) erodes contestability and human agency at scale, or (3) creates disparate impact withou…