Only 25% of Teams Use AI Daily — and IBM's New Study of 2,000 CEOs Explains Why That's the Real Problem in 2026

There is a contradiction at the heart of the organizations that talk most about AI in 2026. According to the IBM 2026 CEO Study, published on May 4 by the IBM Institute for Business Value (IBM IBV) in partnership with Oxford Economics, 86% of CEOs believe their employees have the necessary skills to collaborate with AI. Yet only 25% of the workforce uses AI regularly as part of their daily work. In companies that claim to have AI as a strategic priority, just one in four employees actually uses it.

This gap is not a technology problem — it is a leadership, organizational culture, and work redesign problem. The study, which surveyed 2,000 CEOs across 33 countries and 21 industries between February and April 2026, finds that the organizations overcoming this divide are not the ones that bought more tools, but the ones that redesigned how work gets done before automating it.

Context: three years of acceleration and what has changed

In 2023, the IBM IBV published the report Augmented Work for an Automated, AI-Driven World, which surveyed 3,000 C-suite executives and more than 21,000 workers across 28 countries. At the time, the central conclusion was that 40% of the global workforce would need to be reskilled within three years as a consequence of AI and automation adoption. The World Economic Forum projected that 85 million jobs would be impacted by 2025 and 97 million new roles would be created. Generative AI was the dominant new variable in the conversation.

Three years later, the landscape not only confirms those projections but reveals more complex layers of the same problem. The 2026 study no longer debates whether AI will transform work — that conversation is settled — but instead asks why most organizations have still failed to capture the value they promised to deliver. The answer lies in three dynamics the new report documents with precision: the adoption gap, the C-suite reorganization, and the emerging horizon of autonomous AI decision-making.

The adoption gap nobody wants to admit

The 25% statistic deserves attention because it challenges a narrative that repeats itself in executive presentations worldwide. Organizations announce AI strategies, hire AI leaders, publish press releases about digital transformation — and simultaneously operate with three-quarters of their workforce not using AI consistently in day-to-day work.

The 2026 study reveals that 83% of CEOs surveyed believe AI success depends more on people's adoption than on the technology itself. That number is striking: leadership knows where the problem is and still cannot solve it. The gap is not one of awareness but of execution.

The 2023 IBM IBV study had already identified a similar dynamic, though from a different angle: organizations that cultivated environments where errors during technology adoption were not penalized showed 22% higher revenue growth than their peers among AI users. The conclusion was that the obstacle to adoption was not technical but cultural. In 2026, the same obstacle persists at larger scale, with generative AI far more accessible and still underutilized.

For those leading Customer Experience operations, this statistic has an immediate implication: competitive advantage in CX will not come from having the best tool, but from having teams that actually use it. And getting teams to use AI requires more than platform access. It requires process redesign, role clarity, and active change management.

The C-suite being reinvented in real time

One of the most striking findings of the 2026 study is the speed at which organizational leadership structures have changed. In 2025, 26% of organizations had a Chief AI Officer (CAIO). In 2026, that number jumped to 76%. In twelve months, the role shifted from differentiator to market standard. And companies with a CAIO report 5% higher returns on their AI investments compared to peers without the function.

But the creation of the role is only the visible symbol of a deeper transformation. The study finds that 77% of CEOs see talent leadership and technology leadership roles converging — which in practice means the traditional separation between the CTO (responsible for the technology stack) and the CHRO (responsible for people) is becoming dysfunctional. AI is not an IT question nor an HR question: it is a question of how work is organized, and neither function in isolation appears capable of answering it.

The study also finds that 85% of CEOs say all functional leaders must become technology experts in their own domains. Delegating the technology agenda to the CTO is no longer sufficient: leaders in marketing, finance, operations, and CX need to understand enough about AI to make informed decisions about how it should be applied within their specific contexts.

In parallel, 59% of CEOs expect the CHRO's influence to grow in the coming years, and 79% are decentralizing decision-making as AI expands across the enterprise. This decentralization movement aligns with a trend identified back in 2023, when the IBM IBV showed that the most successful organizations in AI adoption were those giving teams autonomy to define their own performance metrics.

"CEOs delivering real results aren't just deploying AI faster, they're redesigning their organizations."

48% of operational decisions in AI's hands by 2030

The data point that perhaps best defines the horizon the study projects involves expectations around autonomous AI decision-making. Today, roughly 25% of operational decisions are made by AI systems where guardrails apply. The expectation of the CEOs surveyed is that this figure will reach 48% by 2030, four years from now. In other words: within the span of a single presidential term, nearly half of the operational decisions of a typical organization will not pass through direct human judgment.

That number is not science fiction or an optimistic consulting pitch. It is embedded in a trend that any mid-sized operation is already feeling: AI systems already make decisions about dynamic pricing, support ticket prioritization, demand routing, credit approval, and hundreds of daily micro-interactions. What 2030 represents is the formalization and expansion of that existing reality.

For CX operations, this projection has direct implications. The 2023 report indicated that 77% of customer service roles would be augmented by generative AI, not eliminated. In 2026, the question is no longer about role replacement but about who supervises, audits, and calibrates the decisions AI is making. The near-future CX professional does not manage interaction volume: they manage the quality of the autonomous decisions that AI systems make on behalf of the organization.

This connects directly to the finding that 64% of CEOs already say they are comfortable making major strategic decisions based on AI-generated input. The delegation is not on its way. It has already arrived at the top.

Reskilling: from the "40%" of 2023 to the granularity of 2026

In 2023, the estimate was that 40% of the global workforce would need requalification over the following three years — which the IBM IBV translated into 1.4 billion people out of a global workforce of 3.4 billion. That was a large and relatively homogeneous number.

The 2026 study refines that projection with an important distinction between two different types of need. Between 2026 and 2028, 29% of employees will need to be reskilled for roles different from the ones they currently hold, while 53% will need additional development to perform their current role more effectively (upskilling). Combined, 82% of the workforce in the surveyed organizations are at some level of accelerated development need.

The distinction between reskilling and upskilling is not merely semantic: it changes the development strategy entirely. Upskilling can happen within the context of the current role, through modular programs and workflow-embedded learning. Reskilling requires career transition, which puts pressure on internal mobility, skills mapping, and the organization's ability to build bridges between functions that currently do not communicate with each other.

The most recent study finds that organizations that redesigned five core areas (technology, finance, HR, operations, and cross-functional collaboration) are four times more likely to deliver on their business objectives. This is the 2026 equivalent of the 2023 finding on operating model: the organizations that restructure how work functions before injecting technology into the process are the ones that capture real results. Those that overlay automation on existing processes only amplify the inefficiencies they already had.

AI sovereignty as a strategic imperative

A theme that surfaces with considerable force in the 2026 study — and that did not carry the same centrality in 2023 — is AI sovereignty. According to the report, 83% of CEOs agree that AI sovereignty is essential to business strategy. The concept encompasses control over which data feeds an organization's AI systems, where that data is processed, who can audit it, and how AI-generated decisions can be contested or reversed.

This concern emerges from the maturation of adoption: companies that spent the last three years implementing AI tools now realize they depend on infrastructures and platforms over which they have limited control. For CX operations, the point of tension is evident. Customer behavioral data, interaction history, preferences, and complaints are the most sensitive assets of any customer service operation, and how that data is processed by AI systems has direct implications for privacy, compliance, and reputational risk.

What to do: the action agenda implicit in the study

The IBM 2026 CEO Study does not publish an explicit action guide with the same structure as the 2023 report (which laid out four detailed steps), but the data converges on clear priorities that any leader can translate into a practical agenda.

The first is closing the adoption gap with active intervention. The problem is not tool access or leadership conviction — it is the absence of process redesign and a culture that sustains experimentation. Organizations that do not penalize mistakes during adoption capture more value from AI.

The second is accelerating the convergence of talent and technology in leadership. If 77% of CEOs already see these roles converging, those maintaining separate silos between IT and people are operating with an anachronistic structure. The CHRO and the CTO (or CAIO) need to work on the same problems.

The third is preparing a governance model for autonomous decision-making. With 48% of operational decisions projected to be made by AI by 2030, organizations that have not yet defined who supervises, audits, and calibrates these systems are running a risk that has not yet appeared in the P&L but will appear the moment the first wrong autonomous decision reaches a customer.

The fourth — and perhaps the most consequential from a workforce perspective — is distinguishing who needs reskilling from who needs upskilling and building separate pathways for each group. Treating both needs with the same generic training program is a waste of the investment.

Frequently asked questions about the IBM 2026 CEO Study and AI at work