Abstract

Artificial intelligence (AI) governance is developing along different paths in the United States (US) and the European Union (EU). While the EU has developed a horizontal, risk-based framework through the AI Act, the US approach remains distributed across sector-specific federal oversight, State legislation and evolving federal policy. For life sciences organisations, the resulting compliance picture is quite complex. The US Food and Drug Administration (FDA) is developing lifecycle and evidence expectations for AI used in medical devices and drug development, while US States impose their own requirements concerning automated decision-making, healthcare communications, transparency and human review. At the same time, the US is building out its national AI policy and advocating for greater national uniformity.

This article examines the emerging US framework, contrasts it with the EU AI Act, and considers how regulatory teams can build governance systems capable of addressing divergent global requirements without unnecessary duplication of controls, evidence and oversight.