CFO Sarah Friar introduced a scorecard measuring AI value via useful work and cost per successful task. The framework prioritizes dependability and return on compute over vague productivity gains. This shift forces enterprises to quantify actual output. Practitioners can now benchmark OpenAI deployments using concrete financial and operational metrics instead of anecdotal success.