How should boards govern AI risk?

Boards must address a significant governance gap, as many currently lack the necessary AI knowledge or experience to oversee systems that increasingly influence core business strategies [2, 4]. To govern AI risk effectively, boards should move beyond passive delegation, which can lead to abdication of their accountability for outcomes [2]. This involves implementing robust enterprise governance frameworks that establish clear principles, roles, and lifecycle controls to ensure fairness, transparency, safety, and privacy [8]. Effective oversight requires a collaborative approach involving executives, legal teams, developers, and data scientists to share organizational responsibility [3]. Boards must also prioritize AI literacy and ensure that governance principles are translated into practical controls and regular board reporting [4, 10]. Given the rising regulatory pressures in the UK, EU, and US, boards must treat AI governance as a critical compliance priority to mitigate litigation, disclosure, and operational risks [1, 6, 7].