Boards are sleepwalking into the AI era. KPMG’s global risk chief has a survival guide

Corporate governance was built for a deterministic world. AI is probabilistic, fast-moving, and already embedded in your core processes.

You walk into any boardroom in corporate America today and you will see two kinds of execs. 

The first is stood at the head of the table, loudly proclaiming everything they are doing to embrace AI and predicting that it will change the world and everything in it. The second is silent, perhaps buried in papers, perhaps with their head firmly implanted in the sand. 

As is so often the case, the most sensible voices are those somewhere in the middle. 

But make no mistake, the AI revolution has executives walking a tightrope. It has exposed many uncomfortable truths, such as the fact that our oversight models were not fit for purpose, and that corporate governance is falling further behind every day. 

Those that survive this new world, will need to adapt — or die. 

We recently published our AI Governance Principles for Boards with INSEAD to serve as a survival guide for boards navigating these challenging times. We deeply examined five key priorities to provide a practical framework for overseeing AI as it moves deeper into strategy, operations, and the core mechanics of value creation across an enterprise. 

Here’s what we found — and where we think we will go next. 

AI is no longer a fringe technology, so stop treating it like one 

The real question when it comes to AI is no longer about where it can create efficiencies, but what kind of enterprise it is helping to build. Leaders should be putting AI at the centre of every conversation, about growth, competitiveness, capital allocation, risk management, and resilience. Tocreate long-term value creation, start with strategic governance. 

Basic AI fluency is now the bare minimum 

Directors do not need to become software engineers, but they need enough AI fluency to question assumptions, understand dependencies, and see where risk is accumulating. In an environment defined by third-party tools, opaque models, fast-moving vendors, and widening attack surfaces, informed oversight is no longer optional. It is a core board responsibility.

Stop the spread of AI slop to keep humans accountable 

Too much of the AI conversation still defaults to productivity. But the bigger leadership theme at play is how organizations redesign work, preserve human judgment, and keeps people accountable for the decisions that matter. Boards that focus only on automation will miss the more strategic issue: whether the enterprise is building a workforce that can use AI well, question it effectively, and remain responsible for its outcomes.

Trust has never been more important than it is right now 

This is where the tensions within a board’s responsibilities become most visible. Companies are under pressure to move quickly, but trust is far easier to lose than to earn. 

Explainability, fairness, accountability, and transparency are not brakes on innovation. They are what make innovation durable. If trust is treated as a communications issue rather than an operating principle, boards will discover too late that adoption without confidence is not a competitive advantage.

Boardroom work is primed for forensic examination – and rightly so

Governing AI is difficult not only because the technology is changing so quickly, but because it exposes the limits of oversight models built for a bygone era. 

Directors are being asked to govern probabilistic rather than deterministic systems, often in turbulent economic conditions and with uneven levels of AI literacy around the table. That requires more than curiosity. It requires discipline and a willingness to rethink how governance is implemented.

There is still time for boards to shape how AI is adopted before damaging practices become normalized. But that window will not stay open for long. Once AI is embedded across core processes, the cost of poor governance will be measured not only in operational failure, but in reputational damage, lost trust, and missed value.

The board’s role is not to slow AI down for its own sake, nor to cheerlead adoption without scrutiny. It is to make responsible acceleration possible: clear priorities, current risk frameworks, visibility into where AI sits inside critical processes, and confidence that management has the talent, controls, and discipline to challenge outputs before they shape decisions. 

The boards that get this right will do more than manage downside. They will help determine whether AI becomes a source of durable advantage or a case study in failed leadership.

AI will not wait for boardrooms to catch up. The question is whether boards are ready to lead before trust, value, and accountability are lost.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.

This story was originally featured on Fortune.com