Before You Hire a CAIO, Fix the CIO Mandate

Boards are hiring Chief AI Officers at a record pace. Ryan Pehrson, CEO of Pharynos, has seen this movie before—and knows how it ends.

Boards are hiring CAIOs at a pace that should feel familiar, because it is. The Chief Data Officer hit the same adoption inflection point around 2015; within a few years, the CDO had the shortest average tenure in the C-suite, fewer than half of companies called the role successful, and Gartner was predicting that three-quarters would be absorbed back into IT. Now CAIO adoption has jumped from 11% to 26% in just two years, per IBM —a pace the CDO took six years to reach.

Both trends reflect an implicit vote of no confidence in the CIO—a belief that the current technology leader can't handle what's next. But the real problem isn't capability; it's clarity. The CIO mandate was never well-defined, and adding titles doesn't fix that; it compounds it. Split responsibilities produce duplicate workstreams, competing roadmaps, and a coordination tax that drains executive attention. The fix is in the boardroom, not the org chart.

We've Seen This Movie Before

The Chief Data Officer was supposed to be the answer to data-driven transformation. Companies hired them in droves. Then the problems started.

CDO adoption rose from 12% of Fortune 1000 firms in 2012 to 68% by 2018, per Tom Davenport and Randy Bean's research in Harvard Business Review. By 2025, 84% of large firms had one. That's the number boards cite when defending the model.

But stability and success lagged. MIT Sloan Management Review's 2025 analysis found 53.7% of CDOs serve for less than three years, and only 47.6% of organizations call the role very successful. Perhaps more telling: 29% of sitting CDOs believe their own role is transitional and will disappear.

The CDO hasn’t gone extinct. But the role proliferated and destabilized. Tenure instability, role ambiguity, and consolidation pressure tell a clearer story than adoption curves alone.

The CAIO is following the same trajectory, only faster. Same symptoms. Same underlying disease: unclear mandate, fragmented accountability, and no governance architecture to make the role work.

The Coordination Tax

When technology leadership is split across multiple C-suite roles, someone has to coordinate them. Aligning priorities, resolving conflicts, reconciling competing roadmaps—this doesn't happen for free. It consumes executive attention, slows decisions, and creates gaps that people route around. That's the coordination tax.

The research quantifies it. McKinsey found that, in one company, 35% of decisions were duplicated across functions, and generating reports consumed more than 1,000 hours of manual preparation time per month across layers. They also note that cross-functional management processes—planning, budgeting, performance reviews—can consume 40% to 65% of management and overhead time. At one global consumer goods company in their research, more than 60% of decisions and reports were replicated across units. Not because people were inefficient, but because nobody had clear authority to stop them.

In practice, this looks like: The CEO who needs two separate briefings to get one coherent technology picture. Duplicate vendor evaluations. The CFO fielding competing budget requests for overlapping capabilities. Strategic decisions stalling because no one can answer definitively whether the CAIO or CIO owns the outcome.

Meanwhile, shadow IT spending rises—particularly in this AI era—not because people are rebellious, but because they need to get things done and nobody has clear authority to help them.

When Mandates Collide

I noticed this pattern in one of my engagements. A board, under pressure to demonstrate AI "seriousness" to shareholders, created a CAIO role as a peer to the CIO. The hire was credentialed and capable.

But by the time I arrived, the coordination tax was already apparent. I found duplicate workstreams and ambiguous ownership of AI governance functions. The CIO presided directly only over central IT in a utility model, while a confederation of business-unit-aligned IT teams, having only a dotted-line relationship to the CIO, marched to a cacophony of different drumbeats.

The CIO's budget had been informally narrowed without a corresponding reduction in accountability. There were at least two de facto technology strategies running in parallel—theoretically aligned but practically competing. Each generated its own capability and technology roadmaps, vendor conversations, and requests for board attention.

The turning point wasn't a talent upgrade. It was a mandate reset.

The diagnosis was straightforward, if uncomfortable: The CIO’s charter had never been updated when the CAIO role was added. The CIO still owned outcomes that the organizational structure no longer gave him the authority to deliver. The CAIO, operating without a defined boundary, had begun filling the vacuum. It wasn’t out of aggression; it was because someone had to. 

We mapped decision rights explicitly: which decisions belonged to the CIO alone, which to the CAIO alone, and which required joint sign-off. We had a board-level conversation to ratify this model. Having first addressed the reasons parallel technology functions formed in the first place, we collapsed the competing roadmaps into a single consolidated view, with both executives accountable to it. A joint AI Steering Committee replaced the informal negotiations that had been burning everyone’s time.

Once the CAIO was on board, the only reasonable solution was clarity through structure.  These are first-class design issues for any new executive role.

A Simpler Fix

The situation above resulted in a joint steering committee and a ratified decision model. It worked to a degree, but with two leaders, two reporting lines, and two sets of board relationships, the coordination burden was reduced but not eliminated.

There's a simpler solution: don't split the mandate in the first place.

A CIO operating at the right level—sharing ownership of enterprise outcomes, co-leading delivery with business peers, extending accountability beyond operating a tech delivery service—already has the ability to govern AI as an enterprise capability.

The honest question for any board considering a CAIO hire might better be framed: Do we have the right kind of CIO? And, if so, do we need a new title or a new mandate for the leader we already have?

It all comes down to whether you have the right person for the mandate and the moment. Hiring a CAIO to compensate for the wrong kind of CIO is, at best, a workaround. It treats the symptom, but the disease is a CIO operating below the altitude the moment requires. 

The Orchestrator Distinction

The question isn't "CIO vs. CAIO." It's "What kind of leader do we need now?"

It’s about behavior, not titles. The best-performing CIOs operate as orchestrators, not utility managers. Orchestrators share ownership of enterprise outcomes with their CxO peers. Major initiatives have a named business co-owner with shared targets, not just a sponsor. Delivery responsibility lies with multidisciplinary teams in which business and IT staff work together. Accountability for standards, compliance, and risk extends beyond IT.

The utility-manager model, by contrast, keeps technology delivery primarily inside IT. Business leaders sponsor initiatives but don't co-own delivery. Handoffs multiply. Duplicate roadmaps become normal.

Gartner's 2024 survey of 2,457 CIOs reinforced the point: 63% of digital initiatives meet or exceed outcome targets when CIOs adopt the “franchiser” (orchestrator) model, versus 43% under the “operator” or utility manager approach. That 20-point gap isn't a capability gap. It's a decision-rights gap. The difference is in the clarity of the mandate, not the depth of expertise.

If a CIO operates as an orchestrator, AI can be governed and delivered as an enterprise capability without creating a second overlapping mandate. If the CIO operates as a utility manager, boards tend to add specialist titles to force cross-functional traction—then pay the price in executive time and duplicated effort for that overlap.

Setting the Conditions for Success

A peer CAIO and CIO setup will only work under conditions most boards don't establish. Before making the hire, boards must decide:

• Who owns AI governance? Policy, risk framework, controls, monitoring, escalation, and accountability for responsible use

• Who owns AI execution? Platforms, vendor relationships, implementation programs, and delivery accountability

•  How and when do these roles converge? Timeline for integration, conditions for sunsetting or merging the CAIO role, or permanent boundaries with explicit scope separation

Many boards skip this step. They make the hire and expect the new leader to figure out the mandate, territorial lines, and how to collaborate effectively. That's how you end up with overlapping responsibilities and a coordination tax that compounds quarterly.

If the board can't answer these clearly—and in writing—before the hire, it is creating a structural problem. (And if you're being hired into this situation, beware.)

The Hire That Matters

Make no mistake: AI needs leadership attention. The question is whether adding a new C-suite title solves the problem or compounds it.

Companies that outperform on digital initiatives share a trait that has little to do with org charts or titles. They give the right leader a mandate that matches the scope of the problem, and decision rights that stick.

Before your next technology leadership hire, ask the frank question: Are you solving the problem, or creating a new one?

Boards that hire for orchestration get integration. Boards that hire for domain management get silos—then add specialist roles to stitch together what the operating model made difficult.