Without C-suite alignment,

AI performance sputters

AI has no single organizational truth

Source: 2026 AI Impact Survey

Morelli spoke recently with a hospitality industry CIO who was frustrated because people throughout the business kept asking what AI solutions the CIO was going to build while providing no context on the business needs or the problems the AI was supposed to solve.

“To get value out of this, you can’t just ask IT what they’re doing or what agents they’re building,” Morelli said. “There needs to be collaboration across these silos.”

In many organizations, the AI conversation defaults to the technology function. AI implementation, vendor outreach, internal strategy and external advisory all flow through the CIO first. The rest of the C-suite is briefed on AI through the CIO’s or CTO’s viewpoint, while accountability for what AI delivers sits across the entire C-suite. Our survey shows the shortcomings of this approach. The same questions about the same AI deployments produce different readings of where the organization stands.

The technology function examines AI through a build and implementation lens, and from that position it reads its AI program as working and advancing. Technology leaders post the highest workforce readiness and audit confidence figures of any executive group in the survey. The favorable view concentrates further at the top of the function.

The CIO sits at the most favorable end of the most optimistic function in the survey and is the voice briefing AI to the rest of the C-suite. Based on that view, the C-suite builds its strategy.

Operations reads AI from the delivery position. The function runs the workforce absorbing AI, owns AI’s operational outcomes and responds when an AI system fails. From that position, the leadership-staff gap inverts. Where the CIO has a more optimistic view than tech staff, the COO is less optimistic than ops staff.

COOs report a fully implemented AI strategy at 11% against 26% from the broader operations function. The COO sees the AI strategy as less complete because the role is closer to the results the strategy has to produce.

The CIO and the COO are the two executives most directly accountable for AI delivery, and they have completely different views of their organizations. The divergence runs through how they assess readiness, what concerns them about agentic AI and what they believe is causing AI to underperform.

The C-suite cannot align on a fix without aligning on the cause. The misalignment is structural. The four functions are not equally represented in AI conversations, and the executives most operationally accountable for AI performance are the least directly engaged in shaping what it is built to deliver.

The CFO is the only executive structurally required to hold both ends. Finance signs the AI investment that produces the technology function's confidence and owns the outcome operations reports as not yet delivering.

CFOs report full workforce readiness at 10% and audit confidence at 30%, both falling between the technology and operations readings. CFOs run tighter near-term budget growth than the broader finance function, defending the investments already made, while holding a more expansive long-term view of AI's share of enterprise spend. That combination of disciplined near-term control and strategic long-term perspectives gives the CFO standing with both the CIO and the COO.

No single executive can construct the organizational view from inside their own function. Even the CFO cannot produce the four-function view alone. In mid-market organizations, the four executives sit closer together, share leadership cadences more frequently, and overlap in scope more directly than at enterprise scale. Despite the proximity, the divergence emerges faster, with less internal distance to absorb the consequences before they spread.

The C-suite is aligned on direction and divided on mechanism

The C-suite agrees on what AI should accomplish but diverges on the mechanism that produces it. The pattern runs through the survey's questions on investment, priorities and leadership. Risk is where the cost lands most sharply.

Risk dominates the AI investment agenda across the three operational functions, but each function defines risk through its own discipline.

AI investment priorities and business goals

Yet the same executives rank risk and compliance fourth among the functions that would benefit the most from additional focus, enablement or incentives. Operations leads at 43%, followed by Finance at 21%, Customer Experience at 18% and Risk/Compliance at 11%.

The C-suite is concentrating AI investment on risk-related work at higher rates than any other priority and ranking the Risk function last for receiving it. The risk paradox is structural. Risk-driven AI is being embedded across operations, finance and technology, and no executive group is directing investment to centralize accountability for it.

In organizations below roughly $2 billion in revenue, the chief risk officer role often does not exist as a dedicated position at the senior leadership table. Risk accountability is distributed across operations, finance, and technology by default rather than by design, and the leadership team has no infrastructure to verify whether the distribution is producing results.

The divergence on risk is not isolated. Each function’s AI priority stack reveals similar goals for AI, but each has different perspectives on how to deliver AI ROI.

How we can help you

SERVICE

Advisory Services -->

SERVICE

Technology Modernization -->

Ready to talk? We’re ready to listen.

Request a meeting -->

Top AI priorities by function

Finance leaders:

Operations leaders:

Technology leaders:

Source: 2026 AI Impact Survey. Note: Percentages denote the portion of respondents who ranked an issue as a top-3 priority

Functional leaders largely agree that risk mitigation, efficiency and strategic decision support are their top AI priorities. The divergence appears when functional leaders describe what drives AI ROI. CFOs and COOs converge on strategy as the dominant driver (61% and 58%, respectively), but CIOs/CTOs rank strategy lower at 34%.

The divergence sharpens on secondary drivers: CFOs and COOs both identify leadership alignment (14% and 19%, respectively), while the CIO is the only executive to elevate the technology stack (25%). The C-suite agrees on the goals for AI. It fragments on the mechanism to reach them.

AI investment is accelerating ahead of proof

The C-suite is committing more capital to AI in 2026 while conceding the AI it has already built is not yet delivering. Eighty-three percent of finance functions are increasing AI budgets, with AI expected to remain a meaningful but contained share of total enterprise spend over the next 24 months.

Yet the performance evidence runs the other way. Asked to identify up to three factors contributing to AI project underperformance or failure, only 8% of respondents selected “AI is not underperforming.” The remaining respondents named specific failure factors across the AI already deployed. Capital is rising despite the abundance of underperformance.

The C-suite is not retreating from the underperformance signal because the external environment is read through a competitive lens. Operations leaders cite competitor moves as the dominant external force at 53%.

Additional pressure builds as autonomous AI becomes the agreed-upon next frontier across industries, and implementation falls on the technology function. Sixty-five percent of technology leaders would currently permit AI agents to make at least moderate-risk decisions. Five percent would authorize fully autonomous decision-making, including approving financial transactions and executing operational changes without human review.

Autonomous capabilities built on AI that has proven to underdeliver lead to stacking upgrades onto existing problems instead of building on solid foundations, compounding the underperformance while demanding additional investment.

The accountability asymmetry on autonomous AI reveals the most direct consequence of the C-suite misalignment. The functions do not share the same concerns about advancing agentic AI. Twenty-three percent of CIOs/CTOs have no concerns about agentic AI implementation. COOs (54%) and CFOs (42%) are much more concerned about regulatory and compliance uncertainty that agentic AI intensifies, compared with just 20% of CIOs/CTOs. As competitive pressure continues to accelerate capability ahead of proof, the functions accountable for proof absorb the operating exposure.

Closing the distance between authorization and accountability is work no single function can do alone. The next wave of AI accountability will be defined by the next steps the C-suite takes together.

Translation across the C-suite is the work AI accountability requires

No single executive sees the full AI picture from their own position, and decisions made from partial views produce consequences that land across all four functions.

The demand: The C-suite is already asking for the work translation requires. The full sample ranks governance, risk and compliance frameworks as the top external support need at 43%, the highest of nine support areas. Three of four executive groups converge on governance as a foundational need, and all four agree that operations should be the focal point of AI adoption. The disagreement is on whether the workforce is ready to do it, and that is where translation work begins.

The work: Translation operationalizes the ability to explain, measure and defend the AI program and enable performance. That requires four things:

1. A shared definition of AI success across the four executive positions, so each function is reporting against the same outcome.
2. Measurement infrastructure that produces outputs each executive can absorb into their own accountability: technology shows what has been built, operations shows what is being delivered, finance shows what the investment is producing, and the CEO shows how those three views hold together.
3. Governance that distributes the load, instead of defaulting to the technology function.
4. Clear decision rights and accountability.

The C-suite that takes translation seriously will defend AI against the scrutiny that has already arrived. In midmarket organizations, where the executives sit closer and overlap in scope more directly, the conditions to achieve translation are already present. The ability to explain, measure and defend AI decisions will belong to the C-suite that builds translation together.

To learn more about how to enhance AI performance while maintaining strong governance and controls — and how Grant Thornton can help you — visit our AI services webpage.

Contacts: