AI Liability on Boards: Business Judgment Rule Requires Strong Oversight

KPMG Law advises that the Business Judgment Rule can shield board members from personal liability in AI-related decisions only if they act in good faith, are adequately informed, and prioritize the company's interests. The firm warns that delegating AI risk oversight entirely to IT departments is dangerous. A 2026 Optro study of 417 executives and risk, compliance, and security professionals found that 34% of organizations have taken concrete measures.
KPMG Law’s analysis places AI oversight squarely within directors’ duties. The Business Judgment Rule may shield board members only when they can show good-faith conduct, sufficient information, and a company-interest rationale. Outsourcing risk control to IT alone is described as a serious liability exposure, and robust documentation plus monitoring are presented as prerequisites for relying on that protection.
The 2026 Optro survey of 417 governance, audit, risk, compliance, and security professionals across North America and Europe found that 34% had acted after flawed AI-agent decisions. Meanwhile, 30% reported unintended agent actions and 25% control failures. Many organizations grant agents transaction, policy, or user-rights authority, and 46% spend more time checking or correcting AI outputs. Although 96% support organizational adaptation, only 9% have fundamentally redesigned workflows.
Count words: First para: KPMG(1) Law’s2 analysis3 places4 AI5 oversight6 squarely7 within8 directors’9 duties10. The11 Business12 Judgment13 Rule14 may15 shield16 board17 members18 only19 when20 they21 can22 show23 good-faith24 conduct25, sufficient26 information27, and28 a29 company-interest30 rationale31. Outsourcing32 risk33 control34 to35 IT36 alone37 is38 described39 as40 a41 serious42 liability43 exposure44, and45 robust46 documentation47 plus48 monitoring49 are50 presented51 as52 prerequisites53 for54 relying55 on56 that57 protection58. Second: The59 2026? Actually "2026 Optr
Stronger board-level AI oversight could affect investors, employees, customers, and regulators. If directors document and monitor automated systems more carefully, firms may reduce opaque or harmful AI-driven decisions, improving accountability for people subject to those systems. If oversight remains weak or is left mainly to IT, errors in transactions, policies, or user access may go unnoticed longer, potentially harming customers, workers, and shareholders. The debate may also shape how organizations interpret legal duties as AI adoption expands. Count: Stronger1 board-level2 AI3 oversight4 could5 affect6 investors7, employees8, customers9, and10 regulators11. If12 directors13 document14 and15 monitor16 automated17 systems18 more19 carefully20, firms21 may22 reduce23 opaque24 or25 harmful26 AI-driven27 decisions28, improving29 accountability30 for31 people32 subject33 to34 those35 systems36. If37 oversight38 remains39 weak40 or41 is42 left43 mainly44 to45 IT46, errors47 in48 transa