Artificial General Intelligence (AGI) Risks: Should Boards Prepare Now

By Adrian Lawrence FCA, founder of NED Capital · Part of the Board Governance Hub

Few topics divide informed opinion as sharply as artificial general intelligence — AI that could match or exceed human capability across most tasks rather than excelling at narrow ones. Some serious people believe it is close and consequential; others think it remains distant or overstated. For a board, the temptation is to pick a side and plan accordingly. That is a mistake, because the honest position is that no one knows, and a governance approach that depends on a particular forecast being right is fragile by design. The better question is not “when will AGI arrive and how bad will it be?” but “how does a well-run board prepare for something this uncertain?” This piece sets out a durable answer.

Start by Accepting the Uncertainty

The foundation of a sensible board approach is intellectual honesty about what is and is not known. Reasonable, well-informed people genuinely disagree about how soon transformative AI might arrive, how disruptive it would be, and how much of the current discussion is signal versus hype. A board that treats an aggressive timeline as certain will over-invest and distract itself; a board that dismisses the whole subject as science fiction may be caught flat-footed by developments that arrive faster than expected. Neither confident position is justified by the evidence. The mature stance is to hold the uncertainty openly — to treat AGI as a real but unpredictable possibility whose implications for the business should be watched and thought about, without pretending to know its timing or shape. Good governance under deep uncertainty is a familiar discipline; boards handle it routinely with other unpredictable forces, and the same habits apply here.

Focus on What the Board Can Actually Control

A board cannot influence whether or when AGI is achieved — that lies far outside its walls. What it can control are two things much closer to home. The first is the company’s exposure: how developments in AI, up to and including far more capable systems, might affect this particular business — its markets, its cost base, its competitive position, its workforce, its risks. That is a concrete, answerable question, and it will have a very different answer for a software company than for a manufacturer or a professional-services firm. The second is the company’s readiness: whether the board’s own processes — its risk oversight, its horizon-scanning, its ability to take a decision quickly when the ground shifts — are robust enough to respond well to surprise. Directing the board’s attention to exposure and readiness, rather than to speculation about the technology’s trajectory, is what turns an unmanageable abstraction into governable work.

Scenario Thinking, Not Forecasting

Because no one can predict the path of AGI, the useful tool is not a forecast but scenario thinking. Rather than betting on a single future, a board can sketch a small number of plausible ones — from “progress is slower and more incremental than the optimists claim” through to “capability advances faster and more disruptively than expected” — and ask, for each, what it would mean for the business and what early signals would indicate that world is arriving. The value of the exercise is not in getting the future right; it is in rehearsing responses, identifying the indicators worth watching, and surfacing the decisions that would need to be made quickly. A board that has thought through “what would we do if this accelerated?” is far better placed than one encountering the question for the first time in a crisis. Done proportionately — a serious discussion, not a standing obsession — scenario thinking is the single most practical preparation a board can undertake.

Make Sure Someone Genuinely Understands the Technology

A recurring weakness in boards facing any technical subject is that no one round the table truly understands it, leaving the board dependent on management’s framing and unable to challenge it. With something as consequential and fast-moving as advanced AI, that gap is a real governance risk. It does not require every director to become a technologist, but it does require that the board can access genuine, independent understanding — whether through a director with real fluency in the area, a well-briefed committee, or trusted external expertise it can draw on. The point is to be able to ask management the awkward, informed question and to evaluate the answer, rather than accepting either breathless enthusiasm or reflexive dismissal at face value. This capability gap is one of the more common reasons boards appoint a technology-literate non-executive, a theme explored in boardroom AI: are NEDs ready to oversee tech ethics and strategy.

Treat It as a Standing Item, Not a One-Off

Because the picture will keep changing, a single discussion followed by silence is the wrong model. The better approach is to fold AI and frontier-technology developments into the board’s regular risk and horizon-scanning work — a recurring, lightweight item that keeps the subject in view and lets the board adjust its assessment as evidence accumulates, rather than a grand one-off exercise that dates the moment it is finished. This is simply good risk governance applied to a fast-moving external force: keep watching, keep the assessment current, and be ready to escalate attention if the signals warrant it. Framed this way, AGI stops being a special and slightly frightening topic and becomes another item the board oversees with the same disciplines it applies to any material uncertainty. That connects to the broader question of how boards keep proportionate oversight of AI without overreacting, discussed in the role of NEDs in overseeing AI-powered businesses.

Clarify Risk Appetite and Avoid Both Traps

Underneath all of this sits a question of risk appetite: how much exposure to AI-driven disruption — and how much investment in preparing for it — is right for this company. There is no universal answer, and a board’s job is to reach a considered one for its own circumstances rather than to import someone else’s. In doing so it must steer between two failure modes. Complacency — assuming nothing will change materially, or that it is someone else’s problem — leaves the business exposed to a shift it did not see coming. Alarm — treating speculative worst cases as imminent certainties — wastes resources, distracts from present priorities and erodes the board’s credibility. The disciplined middle path is proportion: take the possibility seriously enough to watch it and prepare sensibly, without allowing it to crowd out the real and present work of running the business. A board that can hold that balance is well prepared, whatever AGI turns out to be.

So should boards prepare now? Yes — but preparation means readiness, not prophecy. The board that will handle AGI well is not the one that guessed the timeline correctly; it is the one that stayed honest about the uncertainty, understood its own exposure, kept genuine expertise within reach, thought through its scenarios, and built the habit of watching and adapting. That is durable governance, and it serves a board well regardless of how the technology unfolds. At NED Capital we help boards find non-executive directors with the judgement and, where needed, the technological fluency to lead exactly this kind of measured oversight — each search led personally by Adrian Lawrence FCA, a Fellow of the ICAEW and former listed-company finance director.

NED Capital | Sister practice of FD Capital | ICAEW practising certificate held by Adrian Lawrence FCA.