Why Your VP of Operations Should Hold the AI Strategy Pen

The Question Every Construction Firm Is Quietly Fighting About

AI ownership is genuinely contested at the executive level right now, and it is not settled as "IT's job." For decades the logic was simple. Software ran on servers, servers belonged to IT, so anything with software in it belonged to IT too.

That logic breaks the moment AI starts doing operational work instead of just running on a machine. This is the COO vs CIO AI ownership fight that Harvard Business Review captured¹: the CIO claimed the company's agentic systems as her domain, and the COO countered that managing an autonomous workforce is operations by definition. Several C-suite roles— COO, CIO, CFO, the head of HR— each had a defensible claim. You can already see the same shift in stories like a board that swapped "digital strategy" for an AI strategy and had to ask who, exactly, now owned the new mandate.

"An agentic workforce is the definition of ops." That was the COO's claim in HBR's Fortune 500 scenario.

Here is why agentic AI changes the calculus. It behaves less like software you install and more like a workforce you manage— it takes assignments, makes decisions, and produces work. That is why operations leaders are staking a claim to it.

One caution. HBR documents the debate; it does not declare a winner. The argument that follows, that Operations should hold the pen in construction, is ours, not Harvard's. To answer it for construction specifically, start with what construction's actual problem is.

Construction's Real Problem Is Operational, Not Technical

Construction's defining problem is operational. The hard truth about construction operations is that productivity has barely moved in a generation, and a lack of technology was never the real bottleneck.

According to McKinsey², construction labor productivity grew only about 10% from 2000 to 2022, roughly one-fifth the rate of the overall economy.

Construction productivity grew just one-fifth as fast as the overall economy over two decades. That makes AI an operations-improvement lever, not just an IT project.

The cost of standing still is enormous. McKinsey's analysis projects a cumulative output shortfall of roughly $40 trillion by 2040 if the industry's productivity trend holds³.

The labor shortage compounds all of it. Fewer skilled workers are available to do more work, and that gap has been severe and persistent for years. When you can't add people, the only way to protect schedule and margin is to make the people you have more productive.

What this looks like on a jobsite is familiar:

• Thin margins that a few bad change orders can erase
• Schedules that slip a day at a time until the float is gone
• Rework that eats the profit you booked at bid

Here's the bridge. If the problem is operational, the lever that fixes it is owned operationally. And here's the part the tech-vendor guides skip: every place AI creates value in construction is a place Operations already runs.

AI's Value in Construction Lives Where Operations Already Works

Every workflow where AI creates real value in construction— estimating, scheduling, RFIs, drawing-conflict detection, field productivity— is a workflow Operations already owns. Follow the value, and it leads straight to the operations org chart.

Start with where AI earns its keep. The operational use cases gaining traction⁴ cluster in the field and the project, not the back office:

Look at the right-hand column. The construction VP of Operations owns operational profitability, end-to-end project delivery, quality and safety, vendor relationships, and the workforce— Bridgit describes the role as having "a hand in just about every department"⁵. This is the domain expert in how the firm makes and loses money, and domain expertise paired with AI is where the real gains show up.

The VP of Operations already owns the P&L, the schedule, the crews, and the project delivery that AI changes. Strategy should sit with whoever owns the outcome.

Contractors confirm where they expect the payoff. Dodge Construction Network found that 85% of contractors expect AI to cut time on repetitive tasks and 75% expect it to help them learn from past project data⁶. Operational gains, named by operators.

Adoption is early but moving fast. RICS found about 45% of firms have no AI implementation yet and 34% are piloting⁷, while Dodge found 51% are actively evaluating AI changes and 40% already carry a dedicated AI budget⁶. Early, uneven, accelerating. None of this sidelines IT. It gives IT a sharper, more important job.

The Pen vs. the Plumbing: What IT Actually Owns

IT does not lose when Operations holds the pen. IT owns the plumbing that makes ops-led AI safe and scalable: data governance, security, and system integration. Both roles matter. Both are true.

Take the strongest objection seriously. Without IT and security governance, ops-led AI turns into shadow AI— ungoverned tools that bypass controls, leak data, and create integration chaos. That risk is real, and preventing it is IT's job. Concede the point fully.

Contractors' own concerns prove it. Dodge found that 57% of contractors cite unreliable or inaccurate output as a chief worry and 54% cite data security and privacy⁶. Those concerns sit squarely in IT's domain— exactly the problems a strong data-governance function exists to solve. Part of that mandate is maintaining an AI tool policy that defines what IT controls before crews start wiring their own tools into company data.

So define the division of labor plainly:

Holding the AI strategy pen means owning the roadmap, the priorities, and the outcomes. Owning the plumbing means owning the data, the security, and the integrations. A firm needs both, assigned to different people.

This also answers the "agentic AI is just software, so it belongs to IT" argument. The pen is about strategy and outcomes, not building models. AI governance in construction works best when strategy follows value— and the value is operational. IT makes that strategy safe to run. So if both roles matter, why not just put them on a committee and let them sort it out?

Why One Owner Beats a Committee (Especially Under $100M)

For a $20M–$100M construction firm, a single accountable owner moves faster than a governance committee that meets quarterly.

The committee argument isn't wrong on paper. Cross-functional governance is genuine best practice at enterprise scale, where a Chief AI Officer chairs a standing committee with real budget and staff. But that's not your firm.

Your firm has a VP of Operations and an IT person already stretched across every system in the building. Meet the org you actually have. Name one accountable owner in Operations, give IT a defined partner role, and let them move— and know the difference between a steering committee and a task force for AI before you default to the slower one.

Committee cadence is quarterly. A single owner reviews job-cost outcomes weekly.

The clock matters. Dodge found that 86% of large contractors believe AI will give them a competitive advantage, versus 69% of small and mid-sized firms⁶. The bigger players are already convinced and moving. Mid-market firms that assign ownership to the wrong seat will spend the next two years in meetings while their competitors compound the learning. Here's what holding the pen actually looks like in the first 90 days.

What Ops-Led AI Ownership Looks Like in the First 90 Days

Ops-led AI ownership starts by tying every pilot to a job-cost outcome and giving one operations leader the authority to kill what doesn't move the number. No tech demos for their own sake.

If a pilot can't be traced to a schedule day saved, an RFI cycle shortened, or a margin point protected, it isn't an AI strategy. It's a science experiment.

Here's the first-90-days sequence:

1. Pick one operational pain with a number attached. Estimating turnaround, RFI cycle time, schedule risk, or rework from drawing conflicts— choose the one that costs you the most today.
2. Scope a pilot owned by Operations, aimed at that number. Before it starts, IT defines the data and security guardrails so the pilot runs safe.
3. Review weekly against job-cost outcomes, not feature lists. Did the number move? Keep going. Did it stall? Kill it without ceremony.
4. Sequence the next pilot from what the first one taught. Build the roadmap out of outcomes, not vendor pitches.

This is where most firms stall, because choosing the right first workflow and wiring it to the right metric is genuinely hard. A focused construction AI strategy built around operations— rather than a company-wide overhaul— is usually the fastest path to a result you can measure.

The pattern is deliberately boring. Pick a number, protect it, prove it, repeat. Boring is what compounds.

Frequently Asked Questions

Who should own AI in a construction company?

AI strategy should be owned by operations leadership— typically the VP of Operations or COO— with IT as the enabler. AI's value in construction lives in operational workflows like estimating, scheduling, and project delivery, so the person accountable for those outcomes is the right person to hold the roadmap.

Is AI strategy an IT responsibility?

No. IT owns the infrastructure, security, and data governance that make AI safe and scalable. The AI strategy and roadmap belong with whoever owns the operational outcomes— in construction, that's Operations.

Why does construction need AI in operations?

Construction labor productivity grew only about 10% from 2000 to 2022, roughly one-fifth the rate of the overall economy², while labor is scarce and margins are tight. AI targets exactly those operational pain points, which is why it reads as an operations initiative rather than a back-office IT upgrade.

How many construction firms are using AI?

Adoption is early but accelerating. RICS found about 45% of firms have no AI implementation and 34% are piloting⁷, while Dodge found 40% of contractors already have a dedicated AI budget and 51% are actively evaluating AI changes⁶.

The Person Who Owns the Margin Should Own the Roadmap

The person who owns the margin should own the AI roadmap. In construction, that's your VP of Operations.

IT should own the plumbing. Operations should hold the pen. In construction, that division of labor is the strategy. And it works because people are the answer: AI amplifies an operations leader's judgment; it doesn't get its own department.

Most $20M–$100M firms don't need a Chief AI Officer. They need help getting their operations leader holding the pen well— which is the gap what a fractional AI officer actually does is built to fill. At Dan Cumberland Labs, we help operations leaders drive AI, not IT departments.

References

1. Harvard Business Review (Toby E. Stuart), "Who in the C-Suite Should Own AI?" (2026) — https://hbr.org/2026/03/who-in-the-c-suite-should-own-ai
2. McKinsey & Company, "Delivering on Construction Productivity Is No Longer Optional" (2024) — https://www.mckinsey.com/capabilities/operations/our-insights/delivering-on-construction-productivity-is-no-longer-optional
3. Construction Dive (reporting McKinsey), "Why Construction Productivity Growth Is Lagging" (2025) — https://www.constructiondive.com/news/why-construction-productivity-lags-mckinsey/736082/
4. BuildOps, "Top AI Use Cases to Maximize Construction Efficiency" (2025) — https://buildops.com/resources/ai-use-cases-construction/
5. Bridgit, "VP of Operations in Construction: Essential Skills to Master" (2024) — https://gobridgit.com/blog/vice-president-of-operations-in-construction-skills-to-master/
6. Dodge Construction Network / CMiC (via Construction Dive), "AI Nears 'Tipping Point' in Construction as Contractors Pilot Tech: Survey" (2025) — https://www.constructiondive.com/news/builders-ai-transform-businesses-survey/807555/
7. Royal Institution of Chartered Surveyors (RICS), "Artificial Intelligence in Construction Report 2025" (2025) — https://www.rics.org/news-insights/artificial-intelligence-in-construction-report