AI Superforecasting for FDA Decisions

Cheap, high‑quality AI forecasters could be used to produce probabilistic assessments of trial designs, expected benefits, and approval likelihoods, which regulators and companies would use to make more transparent, incentive‑aligned choices. Embedding such forecasts into procedural rules (e.g., explicit probability thresholds, automated veto windows, or public forecast dashboards) could reduce subjective bargaining, shorten delays, and alter how pharma conducts trials. — If adopted, this would reframe regulatory authority from opaque expert discretion to measurable, probabilistic governance — affecting patient access, drug prices, industry strategy, and legal accountability.