The moment a board starts talking seriously about artificial intelligence, the conversation tends to land in the same place. Someone proposes a Chief AI Officer, a single senior figure to own the technology, set the strategy, and prove the organisation means business. Across boardrooms in London and Europe that appointment has quietly become the accepted marker of ambition. Hire the title, the thinking goes, and transformation will follow.

The instinct is understandable, and it is mostly misdirected. The pressure to install AI leadership is real, but it is aimed at the wrong problem. What we see inside live searches tells a different story. Briefs that open with a request for an AI leader almost always shift, within weeks, towards something far less glamorous: someone who can fix the data and quickly. The appetite for AI is not the constraint. The foundation beneath it is. Companies are trying to hire their way to an end state before they have built a starting point, and the distance between those two things is where most AI programmes quietly stall.

Start with demand

Start with the demand signal, because it is the clearest tell. Clients come to us convinced they need AI leadership, and then the conversation turns practical. Their data is fragmented across systems that were never designed to talk to each other, inconsistent in how it is captured, and largely unusable in its current state. Almost every brief that begins with AI ends up centered on interim data-focused roles instead. Of the Interim AI client briefs I took over H1, 90% ended up requiring a Chief Data Officer, Head of Data & AI, Data Consultant, or Head of Data & Analytics. This is not confined to one sector. It shows up across financial services, industrials, healthcare, and consumer businesses alike, which tells us it is a structural issue rather than an isolated one. Organisations are solving for the destination when they have not yet mapped the road, and the real work sits upstream in the plumbing nobody wants to discuss.

There was also a widespread expectation that AI would create a permanent new seat at the top table. In practice, that seat is not materialising at scale. Where AI leadership does land, it is being absorbed into existing mandates, into the Chief Data Officer's remit, or into the CIO and CTO functions that already own the underlying systems. In private equity–backed companies the resistance is sharper, because few sponsors will fund another C-suite layer without a clear return. Beneath that lies a second problem. The pool of leaders with genuine applied AI experience, as opposed to conference-stage theory, is thin. Appoint a Chief AI Officer too early and you tend to get role ambiguity, a slow start, and a mandate that drifts from what the business actually needs.

The ROI on AI

The return on AI investment is the third pressure point, and it exposes the same root cause. Leaders are being pushed to invest, often without a defined use case, and the results are hard to quantify. MIT's Project NANDA found that 95% of enterprise generative AI pilots deliver no measurable impact on the profit and loss account, a figure that says less about the technology than about the conditions it is dropped into (https://www.tomshardware.com/tech-industry/artificial-intelligence/95-percent-of-generative-ai-implementations-in-enterprise-have-no-measurable-impact-on-p-and-l-says-mit-flawed-integration-key-reason-why-ai-projects-underperform). When systems are disconnected, governance is weak, and data is poorly structured, even a capable model has nothing solid to work with. The pattern repeats itself. A company pushes hard on AI strategy, hits the data wall, and steps back to invest in cleanup, governance, and foundational modernisation.

What if you really do need an AI leader?

Some organisations genuinely do need a dedicated AI leader now, and a handful already have the clean, connected data to justify one. Technology firms and a few digital-native businesses sit in exactly that position, and for them the appointment makes sense. That exception proves the wider point rather than undermining it. Those companies earned the right to hire an AI leader precisely because they fixed the foundation first. For the majority still wrestling with fragmented systems, the sequence matters. Hire the strategist before the data is ready, and you have bought ambition without the means to act on it.

Don’t just do something

The pressure to "do something with AI" is not going to ease, and nobody is suggesting you do nothing. The mistake is moving quickly in the wrong direction, because that carries more risk than moving deliberately in the right one. Most organisations are not blocked by a shortage of AI ambition. They are blocked by data that cannot yet support it. The companies that pull ahead over the next 24 months will not be the ones that rushed to appoint an AI leader. They will be the ones that built the foundation first. Before you hire for AI, ask the harder question: is your data ready to support it? If the honest answer is no, your next hire is not a Chief AI Officer. It is the interim leader who can fix what sits beneath one.

‍