Who Should Own AI in the C-Suite? Why the Wrong Question Is Killing Corporate AI Strategy

A strategic breakdown of the Chief AI Officer (CAIO) role, jurisdictional competition among executives, and the governance framework actually needed for enterprise AI deployment.

The Boardroom Fight Nobody Prepared For

In January 2026, the CEO of a Fortune 500 insurer locked the doors of a conference room and asked a single question: Who owns our AI strategy?

The Chief Information Officer (CIO) claimed AI was infrastructure. The Chief Operating Officer (COO) argued AI agents execute business workflows. The Chief Financial Officer (CFO) pointed to AI-driven underwriting decisions hitting the P&L directly. The Chief Risk Officer (CRO) flagged autonomous decision-making as a material risk exposure. The Chief Human Resources Officer (CHRO) insisted AI agents functionally are employees. The Chief Data Officer (CDO) closed the loop: without data rights and quality, none of the above mattered.

Six executives. Six legitimate claims. Zero resolution.

Three months later, the company drafted a job description for a Chief AI Officer (CAIO)—an external hire tasked with solving a problem the C-suite could not.

This is not a parable. It is the opening of a 2026 Harvard Business Review article by Toby E. Stuart, professor at UC Berkeley’s Haas School of Business. The scenario captures a structural pain point thousands of enterprises are experiencing but few can articulate: AI does not fit into existing organizational boundaries.

Why AI Is Not the Next ERP or Cloud

Corporate veterans will recognize the pattern. When ERP systems arrived, business units and IT clashed until governance settled under the CIO. When cloud computing emerged, business leaders bypassed IT with corporate credit cards until spiraling costs and security gaps forced the creation of the Chief Data Officer (CDO) role.

Past technological waves were ultimately absorbed into technical hierarchies. The default assumption today is that AI will follow the same path: give it to the technology function and move on.

That assumption is wrong.

An autonomous claims-processing AI agent is simultaneously:

• A technology system (CIO/CTO)
• A workflow operator (COO)
• A profit-and-loss decision maker (CFO)
• A regulatory and litigation risk (CRO / General Counsel)
• A functional employee requiring onboarding and collaboration (CHRO)
• A data-dependent asset requiring governance (CDO)

When a single asset triggers legitimate ownership claims from six C-suite functions, the organization is no longer debating tool ownership. It is debating who controls the steering wheel of the company itself.

Jurisdictional Competition: The Sociology of C-Suite Conflict

Stuart’s article introduces a powerful lens from sociologist Andrew Abbott’s 1988 book The System of Professions: jurisdictional competition.

Abbott’s theory states that professional groups perpetually contest boundaries over who has legitimate authority over specific domains. In corporate terms, this is the mechanics of “turf wars.”

Abbott’s critical insight: major technological disruptions redraw these boundaries. Supply chain, operations, R&D, IT, HR, finance, and marketing coexist with clear borders during stable periods. AI collapses those borders because it simultaneously creates new work (prompt engineers, agent trainers, deployment specialists) and rewrites existing power dynamics.

The fight over AI ownership is not petty politics. It is structural. And it is only beginning.

The Four Default Solutions—And Why They Fail

Organizations facing this ambiguity typically default to one of four paths:

1. Give it to the most capable leader — Overloads one function with cross-domain scope it cannot execute.
2. Give it to the loudest voice — Decisions become political rather than strategic.
3. Assign a committee — Temporary structures move too slowly for AI’s velocity and dissolve into competing interests.
4. Attach it to the existing budget owner — Fragments AI strategy across siloed financial controls.

Stuart’s argument is direct: AI is too large, too fast, and too broad for any of these patches.

The Reframe: From “Who Owns AI?” to “Which AI Decisions Belong to Whom?”

The central strategic shift proposed in the research is a change in framing.

Stop asking: Who has full decision rights over AI?Start asking: Which AI decisions should be owned by which leader?

The first question is zero-sum. The second is decomposable.

Governing an AI agent requires distinct decision types:

• Strategic objective: What business problem does this agent solve? → COO
• Technical architecture: How is it built, secured, and integrated? → CIO / CTO
• Data governance: What data can it access, and at what quality? → CDO
• Risk threshold: What are the escalation triggers and kill switches? → CRO / Legal
• Workforce design: How do humans and agents collaborate? → CHRO
• Capital allocation: What is the investment, cost structure, and ROI? → CFO

These are different categories of decisions with different natural homes. The goal is not consolidation under one crown. It is explicit partitioning with clear accountability.

What the Chief AI Officer (CAIO) Actually Does

This brings us to the CAIO role itself. Many companies create the position out of confusion: if no existing executive can own AI, hire someone who can.

Stuart warns: A new role succeeds only when it has a coherent jurisdiction. If the CAIO is defined as “the person who owns AI,” the position is designed to fail. AI touches everything; therefore, a CAIO who tries to control everything controls nothing.

The correct mandate is coordination. Specifically, the CAIO should own the AI Decision Rights Map:

• Create the map: Identify which AI decisions belong to which functions.
• Maintain the map: Update boundaries as technology and business evolve.
• Identify gaps: Surface overlaps, blank spaces, and conflicts.
• Drive clarity: Work with the CEO and functional leaders to codify decision rights, veto powers, consequence ownership, and review cadences.

The CAIO is not an AI overlord. The CAIO is the cartographer of AI governance.

This is a genuinely new jurisdiction. No previous executive was formally responsible for “autonomous system cross-functional coordination.” As AI penetrates deeper into operations, this coordination function becomes more critical than any single technical deployment.

The market is responding. CAIO roles are proliferating rapidly, with median compensation in the U.S. approaching $1.6 million annually and top packages reaching $3.5 million. In China, institutions recently published the Beijing Consensus on Chief Artificial Intelligence Officers, systematizing the CAIO function.

Yet in every market, the pattern is consistent: CAIOs are cross-functional coordination hubs, not AI commanders-in-chief.

The Dynamic Reality: Decision Rights Must Be Redrawn Repeatedly

There is no permanent answer to who should own which AI decision.

AI’s impact is too wide, and its evolution is too fast. Boundaries drawn today will likely be insufficient within quarters, not years. Jurisdictional competition will persist. Newly formed balances will be broken by the next wave of agentic capabilities.

The new leadership discipline of the AI era is not answering “Who owns AI?” once. It is building the organizational muscle to redraw the ownership map every time technology outpaces the current structure.

Strategic Takeaways for the AI Era

Whether you are in the C-suite or building toward it, three organizational trajectories are becoming clear:

1. Hybrid fluency is the new power. Leaders who can bridge business context and AI capability will disproportionately rise. Functional silos are losing their protective walls.
2. Organizational structure is becoming a competitive variable. Enterprises that treat AI governance as a static reporting-line question will lag. Enterprises that treat it as a dynamic decision-rights architecture will adapt faster.
3. Action velocity and learning velocity are converging. The half-life of technical obsolescence is collapsing. The ability to learn, re-skill, and re-draw boundaries is becoming more valuable than any fixed domain expertise.

The One Question to Ask in Your Next Strategy Meeting

The next time you hear a colleague ask, “Who is in charge of AI here?”—interrupt them.

Ask instead: “Which AI decisions are we making, and who is explicitly accountable for each one?”

Watch who tries to redraw the map. Watch who tries to grab the map. The difference tells you everything about who is built for this era.

Further Reading:Stuart, Toby E. “Who in the C-Suite Should Own AI?” Harvard Business Review, 2026. (For the full theoretical grounding on Andrew Abbott’s jurisdictional competition framework and enterprise AI governance models.)