The Board Does Not Need Another AI Warning. It Needs a Leader.

Boards govern what they are given. Give them something worth governing.

Paul Sala   |   Co-Founder, Ibex Ascent   |   May 2026  

AI conversations are happening everywhere in most large organisations right now, and that is part of the problem.

Technology teams are raising legitimate concerns about cyber risk, data readiness, infrastructure cost, and change fatigue. Business operations leaders are working through what an AI-affected operating model means for their people and processes. Commercial and product functions are energised by early pilots and whiteboard possibilities. And every week, another headline announces a capability leap or a projection about workforce displacement.

Boards are receiving all of this simultaneously. What they are not receiving is a coherent ask.

That is a management problem. And it is entirely solvable.

Only 4% of organisations are currently creating substantial value from AI, and just 22% have moved beyond proof of concept, according to BCG. Only 39% of organisations report any enterprise-level EBIT impact at all. Those figures are not evidence of resistant boards or cautious executives. They reflect roughly where the adoption curve sits when the internal case has not yet been made clearly enough to earn a mandate. Boards know how to govern disruptive technology. They have done it before. What they need from their management teams is clarity, not a catalogue of perspectives.

The conversation that needs to stop

There is a version of the AI briefing that has become so familiar it has stopped producing decisions. It goes something like this: AI is not just a technology change, it is an operating model change. It is not about replacing people, it is about business value. Every AI initiative surfaces a data problem. The workforce implications require careful management. The risk landscape is evolving.

All of that is true. Boards already know it. Repeating it does not constitute a plan, and arriving at a board meeting with a well-organised summary of complexity is not the same as arriving with a recommendation.

The organisations pulling ahead are not having a more sophisticated version of that conversation. McKinsey's research shows that organisations redesigning end-to-end workflows see the greatest EBIT impact, and that for every dollar spent on technology, roughly five dollars should be invested in people and adoption. They got there because someone in management put a coherent case together and asked the board to approve a way forward, not a way of thinking about it.

What boards are actually waiting for

Boards are not looking for an AI strategy document. They are not looking for a technology roadmap or a market benchmarking exercise. They are looking for the same thing they look for when any significant business change lands on the agenda: a credible recommendation from the people they hired to make them.

What are we trying to achieve? What are we prepared to invest and change to get there? How will we govern it? What will come back to us for decision, and when? And critically: what is the signal that something is not working, and what do we do when we see it?

These are the same questions a board would ask about any significant technology-enabled business change. The transformation leader's job is to answer them clearly, not to educate the board about AI.

That distinction matters. Presentations that lead with AI complexity, technical architecture, or transformation risk tend to produce caution. Presentations that lead with business outcomes, investment logic, governance design, and scale criteria tend to produce mandates.

The case in practical terms

A board-ready AI programme does not need to resolve every question before asking for commitment. It needs to demonstrate that management has a credible method for working through questions as they arise, and that the board will retain visibility and control throughout.

The structure is straightforward.

Start with posture. How ambitious does the organisation want to be, and over what timeframe? A selective approach concentrates on a small number of high-value use cases with tight approval gates. An aggressive approach accepts more uncertainty in exchange for faster learning. A transformational approach is prepared to redesign core workflows and accountability structures where evidence supports it. The board sets the posture.

Management delivers within it.

Move to investment logic. What is the organisation prepared to spend to learn, prove value, and build capability? Frame this as a staged commitment: an initial envelope for the learning phase, with scale investment contingent on evidence. This is familiar ground for any board that has governed a major technology programme.

Define risk appetite. Not every risk scenario needs resolution upfront. The board sets the tolerance: what is acceptable without further approval, what requires explicit sign-off, and what is outside the boundary regardless of potential value. That becomes the operating framework.

Establish scale criteria. What evidence is required before the organisation invests more heavily? Agree in advance what a successful adoption trial looks like, what a credible value measure is, and what constitutes a genuine reason to stop rather than adapt.

Finally, the rip cord. Define the conditions under which the programme escalates to the board, pauses, or stops. What risk threshold triggers a review? What benefit shortfall prompts a redirect? What adoption failure warrants a different approach? A management team that has thought this through is not being pessimistic. It is being professional.

What the programme delivers

The programme itself follows a logic boards will recognise from every major technology wave before this one.

Select priority use cases based on business value, feasibility, risk, and learning potential. Run adoption trials rather than technology pilots. The distinction matters: a technology pilot tests whether the AI works. An adoption trial tests whether the organisation can use it safely, repeatedly, and with measurable benefit in real work. Each trial produces a clear decision to scale, adapt, pause, or stop. All four are legitimate outcomes if the learning is honest.

As use cases mature, they surface the specific data, process, governance, and workforce requirements needed to go further. Those become funded workstreams when the evidence justifies the investment. Not abstract preconditions imposed before anything starts, but concrete decisions that emerge from real work. The operating model implications, the data challenges, the workforce questions: they do not disappear. They become manageable, sequenced, and owned.

Over time the programme builds reusable capability. Governance patterns. Data access models. Human review protocols. Benefit measurement disciplines. Each subsequent wave moves faster because the organisation is genuinely learning rather than repeatedly starting from scratch.

The transformation leader's moment

BCG's 2025 research found that only 5% of companies qualify as genuinely future-built for AI, driving substantial value through innovation, while 35% are scaling AI and beginning to generate value. Those numbers will move. The technology is not going to become less capable or less relevant. The organisations building structured approaches now will have a learning advantage that compounds.

This is, straightforwardly, what transformation leaders have been hired to do. Not to facilitate the AI conversation. Not to coordinate perspectives from technology, operations, HR, and risk into a balanced summary. But to synthesise all of that into a coherent recommendation the board can act on, then deliver the programme with the discipline and transparency that keeps confidence high.

The organisations that will look back on 2025 and 2026 as the period when they built a genuine AI advantage are not going to be the ones that had the most pilots or the most ambitious vision statements. They are going to be the ones where someone in management got the board message together, presented a practical plan with a clear value proposition, and had the professional confidence to say: here is how we move forward, here is how we will know it is working, and here is what we will do if it is not.

Boards understand disruptive technology. They have navigated it before. Give them the plan.