The era of AI governance living comfortably in the management layer is over. For years, boards were briefed, questions were asked, and executives returned to their teams to figure it out. That arrangement has collapsed. In the space of a few months, the conversation has shifted from whether boards should engage with AI to: how, how often, and with what level of personal accountability attached to the answer.
David Fapohunda, a business strategy, operations, and AI enablement executive with over two decades of leading large-scale AI and digital transformations across global organizations, has watched this shift accelerate, and his assessment of where boards now stand is unsparing. “AI governance is now a fiduciary discipline,” Fapohunda states. “The boards that treat it that way will be protected. The boards that do not will be visible to investors, to regulators, and to the courts.”
The Gap Is No Longer a Weakness. It Is a Governance Risk
In April, KPMG and INSEAD released a global set of AI governance principles. The message was not subtle. Boards are now expected to demonstrate active oversight of AI strategy, AI security, workforce transformation, and the trustworthiness of every model the enterprise deploys. The KPMG Global AI Pulse Survey found that nearly three-quarters of boards are perceived to have only moderate or limited AI expertise. That is a governance liability that exists right now.
Fapohunda is precise about what closing that gap actually requires. Directors do not need to become technologists. They need genuine fluency – enough to ask probing questions and, critically, to recognize when the answers coming back from management are insufficient. A board that cannot identify an inadequate answer cannot provide meaningful oversight. The briefings mean nothing if the people receiving them cannot evaluate what they are hearing.
Investors Are No Longer Watching. They Are Scoring
The external pressure has become concrete and personal. Glass Lewis signaled in its 2026 guidance that where insufficient AI oversight results in material harm to stakeholders, it will identify which directors or committees were responsible and may recommend voting against them.
Morgan Stanley and BlackRock’s analysis now links AI governance maturity directly to valuation. Opaque or unmonitored AI deployment is no longer an operational footnote; it is a market signal, a litigation exposure, and a named accountability question for individual directors sitting around the board table. The organizations with mature AI governance are being rewarded. The ones without it are being flagged. That is no longer a future risk to be managed. It is the present reality being priced into how companies are perceived, valued, and held accountable.
Audit Committees Are Moving. Boards Need to Move With Them
The most practical shift is structural, and it is happening fast. Audit committees are absorbing AI oversight responsibility at speed and boards are running tabletop exercises on AI failures the same way they run them on cyber incidents and financial stress scenarios. Moreover, management is being asked to produce not a slide deck about AI strategy, but a coherent, enterprise-wide account of how AI behaves today, how it will behave tomorrow, and how it performs when things go wrong.
The boards asking those questions and holding management accountable for the answers are building a governance posture that will withstand regulatory scrutiny, investor pressure, and the legal exposure that follows when it does not. The boards that have not yet made that shift are not invisible. They are simply leaving it to someone else to name the gap first.
Follow David Fapohunda on LinkedIn for more insights on AI governance, board accountability, and building the enterprise frameworks that make AI a strategic asset rather than an unmanaged risk.
