It is a business question first
Most companies I speak with have already started with AI. There is a working group, a few pilots, a vendor or two, and a budget line inside the technology plan. None of that is wrong. The trouble is where it sits. When AI lives inside the technology agenda, the questions it answers are technology questions: which tools, which platform, which data.
The questions that decide whether it pays off are different. Which clients will we serve better? Which costs should fall, and by how much? What will our competitors be able to do in two years that we cannot? Where should we stop spending money because a machine will soon do the work for a fraction of the cost?
Those are strategy questions, and in a company between $500 million and $2 billion in revenue, strategy questions belong to the CEO. Your CIO is essential to the answer. The CIO should not have to own the question alone.
What happens when it is delegated
Delegation looks efficient. The CEO names a sponsor, the sponsor forms a team, and the team reports back each quarter. In practice, three things tend to happen.
- Pilots multiply. Every function proposes its own use case, each one reasonable on its own, and nobody has the standing to say which three matter more than the other twenty.
- Funding stays small. Projects are sized to fit inside existing budgets, so nothing is large enough to change how a business line competes.
- The operating model never moves. AI gets added to today's process instead of changing who does what, so the savings show up on a slide and not in the P&L.
None of this is a failure of the technology team. It is what happens when a decision that cuts across the whole company has an owner in one function. The sponsor can recommend, but cannot reallocate budget from sales to operations, or tell a business unit president that their favorite project is not a priority this year.
AI decisions are capital allocation decisions. They belong where capital allocation already lives.
What CEO ownership looks like
Owning AI strategy does not mean the CEO becomes the most technical person in the room. You do not need to understand how a model is trained any more than you need to understand how your ERP stores a journal entry. It means doing the few things only a CEO can do.
Set the ambition
Say plainly what AI should change about the business: growth in a segment, a cost position, the speed of a core process.
Choose, and say no
Pick a small number of priorities and stop the rest, including projects with loud sponsors.
Assign business owners
Each priority is owned by the leader whose results it changes, not by the team that builds it.
Hold the review
Put AI progress on the executive agenda on a fixed rhythm, measured in business terms.
The last one matters more than it looks. What the CEO reviews is what the organization believes is important. If AI shows up once a quarter as a technology update, people will treat it as one. If it shows up every month next to revenue and margin, with a business leader explaining what moved, people will treat it as part of how the company runs.
Where the CIO fits
Moving ownership to the CEO raises the CIO's role rather than reducing it. In the companies that do this well, the CIO is the person who tells the executive team what is feasible, what it will cost, what data the business actually has, and which risks need managing. That is a seat at the strategy table, not a request queue.
The practical shift is in the order of the conversation. The executive team agrees on the business outcomes first, then asks the CIO how to get there. The CIO brings options and tradeoffs. The decision is shared, and so is the accountability.
A practical first step
If this sounds like a reorganization, it is not. Most of the work is making decisions that are already overdue. Start by putting AI on the agenda of your next executive offsite as a business topic, with the CIO presenting options and each business leader answering one question: where would AI move my numbers?
From there, the next step is choosing your first three priorities, which deserves its own discussion. The point for now is simpler. AI will change how companies in tech-enabled services compete. The CEO should decide how it changes yours, before the market decides for you.
