For structured extraction tasks (identifying specific clause types, extracting defined terms, finding dates), AI achieves 95–99% accuracy. For nuanced risk assessment — whether a clause is "market standard" or "high risk" for your context — accuracy depends heavily on how well the system is calibrated to your preferred positions and risk tolerance. We use your historical redlines to calibrate.