AI SAFETY & ETHICS

The Denominator Problem in AI Governance

Tech Policy Press

Michael A. Santoro argues that counting AI harms without measuring opportunities for harm makes the data uninterpretable - and undermines effective AI governance.