The number that stands out: 4 percent
Gartner surveyed 2,002 CIOs and technology executives for its 2027 CIO and Technology Executive Agenda report, released October 19 at the firm's IT Symposium/Xpo in Orlando, and the headline finding describes a wide gap between enthusiasm for AI and actual willingness to restructure how work gets done around it. Only 4 percent of CIOs report having concrete, documented plans to redesign roles for a workforce that blends human employees and AI agents working side by side.
That is a strikingly small number given how much budget and executive attention agentic AI has absorbed across the industry over the past year, with boards pushing for faster deployment at nearly every company Gartner surveys. Most organizations, by Gartner's own data, are instead deploying agents directly into existing workflows and existing job definitions, rather than pausing to rethink what those workflows and job definitions should actually look like once agents are doing a meaningful share of the work.
The payoff the 4 percent are capturing
The organizations that have actually done this structural work are not seeing marginal, hard-to-measure gains for the effort involved. Gartner found that enterprises redesigning roles for human-agent collaboration are 2.4 times more likely to be classified as top AI performers in the same survey, and considerably more likely to have CEOs who report that AI is delivering real, measurable value rather than the diffuse promise most leadership teams have been asked to take on faith for the past two years.
That gap is some of the clearest evidence yet that the real difference between organizations getting value from AI and organizations merely spending heavily on it has little to do with model selection or vendor choice, the two variables most procurement conversations still fixate on. The deciding factor appears to be whether the business actually redesigned how work flows through the organization once agents became part of the workforce, rather than simply bolting new agents onto an org chart that never changed.
Why 13 percent is the number CIOs should worry about
Despite years of sustained AI investment across the enterprise sector, only 13 percent of CIOs report significant or transformational value from that spending in Gartner's survey. That is a genuinely sobering number set against how much capital has already been committed to AI infrastructure, licensing, and integration work industry-wide, and it should temper any comfortable narrative that AI adoption automatically compounds into AI value simply by accumulating more time and more deployment.
At the same time, 58 percent of leaders report rising pressure from within their own organizations to show concrete cost improvements from AI investments already made. CIOs are effectively being squeezed from both directions at once, told to keep spending to stay competitive while simultaneously being asked to prove the prior spending was worth it, and doing so with relatively little hard, concrete value to point to as evidence just yet.
Pressure from the top, inertia in the middle
Gartner VP analyst Alicia Mullery frames the dynamic as a leadership-driven shift rather than something bubbling up organically from frontline teams experimenting on their own initiative. Pressure is coming from the CEO, boards, and leadership, she says, to move past unconditional AI expansion and start demonstrating actual return on everything already committed to agentic tooling and infrastructure. That pressure from the top has not yet translated, at most organizations she tracks, into the kind of structural workforce redesign that Gartner's own data shows is what actually produces results worth reporting to a board.
Mullery's broader read on the moment is that expectations at the leadership level have shifted well ahead of actual behavior on the ground at most of the companies Gartner tracks closely. I think we've passed the time of endless scale, she says, describing a move away from land-grab style AI deployment toward a considerably more disciplined phase where leadership now wants hard evidence, not simply rising adoption numbers, before authorizing the next round of investment into agentic tooling across the business.
What workforce redesign actually requires
Redesigning roles for a genuine human-agent workforce is meaningfully harder than simply approving another agent deployment, because it touches job descriptions, performance metrics, and reporting lines directly, rather than just adding another line to a tooling budget. It requires deciding which decisions an agent can reasonably make unsupervised, which decisions still require human sign-off, and how a worker's actual role and expectations change once part of their prior workload has been automated rather than merely augmented alongside it.
That kind of work is slower and considerably more politically sensitive than a straightforward procurement decision, which likely explains why only 4 percent of CIOs have gotten there despite near-universal AI experimentation happening everywhere else in the business. It requires genuine buy-in from HR, legal, and line-of-business leaders in a way that simply buying another AI tool never does, and most organizations have not yet built the cross-functional process required to do this work properly.
The roadmap call
The practical takeaway from Gartner's data is that CIOs should stop treating workforce redesign as a phase two problem that quietly follows AI deployment at some unspecified later date, and start treating it as the thing that actually determines whether deployment produces real value at all. A 2.4 times performance gap between organizations that do this work and organizations that do not is simply too large for any serious CIO to file away as a nice-to-have initiative for next year's roadmap.
A reasonable first step, short of a sweeping organizational mandate, is picking one workflow where agents are already deployed and formally redesigning the roles, metrics, and decision rights built around it, treating that as a deliberate pilot for the broader change rather than an afterthought. The 96 percent of CIOs still without a documented plan are deploying AI at scale already. The work they have left undone is changing the roles and incentives around that deployment, and Gartner's numbers say that gap is exactly where the value is actually sitting.
