Across industries, the push toward AI adoption has moved from experimentation to expectation. Budgets are allocated. Tools are deployed. Roadmaps are drafted. On paper, progress looks strong.
But a closer look at the data tells a more complicated story.
In a recent survey of more than 500 senior leaders, 93 percent say they actively encourage their teams to use AI, and 82 percent report regular use across their teams.
At the same time, only 27 to 28 percent are applying AI to more strategic work like scenario planning, organizational design or financial modeling. The gap between intent and execution is clear — and widening.
This is the AI competency gap: the distance between how ready leaders believe their organizations are to operationalize AI — and how ready they actually are.
For CLOs and learning leaders, this gap shows up in stalled initiatives, uneven adoption and teams waiting for clearer direction. And increasingly, it traces back to leadership itself.
The leadership bottleneck no one planned for
One of the most consistent signals in the data is where capability breaks down.
Vice presidents, those closest to translating executive vision into operational reality, are falling behind.
Only 73 percent of VPs have completed AI training, compared with 88 percent of directors. And when it comes to leadership-specific AI training, the gap widens further: Just 55 percent of VPs have participated in the past year, versus 80 percent of directors.
That disparity carries through to real competencies.
While 68 percent of leaders overall say they understand how to use AI without compromising company data, only 58 percent of VPs report the same confidence. The pattern repeats across areas like vendor decision-making, workflow design and team enablement.
The result is a structural weak point. Strategy may be set at the top. And execution may be happening through teams. But the layer responsible for connecting the two is often the least prepared to do so.
“Organizations don’t struggle with AI because they lack tools,” says Daniele Grassi, CEO of General Assembly. “They struggle because leadership capability hasn’t caught up to the pace of investment.”
That mismatch creates friction in places CLOs know well: initiatives that launch but don’t scale, pilots that fail to translate into practice and teams that default to old workflows despite new technology.
When AI stays tactical, transformation stalls
Even where adoption is high, usage patterns reveal another constraint.
Most leaders are engaging with AI, but primarily at a surface level.
Sixty-nine percent use it for search. Sixty-eight percent use it for summarization. Fifty-eight percent use it for drafting communications. These are useful, but they’re not transformative.
More strategic applications like scenario planning, organizational design and resource allocation remain far less common, hovering around 27 to 32 percent.
This matters because enterprise adoption depends on how leaders use AI, not just whether they use it.
If leaders treat AI as a productivity tool, their teams will follow. If they use it to rethink decisions, redesign workflows and challenge assumptions, the organization begins to shift.
Right now, most organizations are stuck at the surface level. And while subtle, the cost is significant.
Teams experiment without clear direction. Use cases remain isolated. Momentum slows. And in some cases, initiatives are rolled back altogether. A quarter of leaders report scaling back AI efforts in the past year, citing issues ranging from data readiness to lack of skills.
“AI fluency isn’t about knowing the tools, it’s about knowing where and how to apply them to real business problems,” says Nick Goldberg, CEO of EZRA. “That’s a leadership capability, not just a technical one.”
Until that capability is effectively built, transformation efforts will continue to remain stuck.
Capability is the differentiator
There is, however, a clear signal in the data around what’s working.
Leaders who’ve participated in structured, leadership-specific AI training consistently outperform their peers. They’re more confident in their skills, more likely to redesign workflows, more likely to have teams actively using AI, and more likely to apply it in complex, strategic contexts.
One example: 96 percent of leaders who’ve completed leadership AI training report regular team use, compared with lower rates overall. Eighty-eight percent say they understand how to use AI tools without compromising data, versus 68 percent across the broader group. They’re also significantly more likely to evaluate AI use in performance reviews and establish clear standards for what “good” looks like.
In other words, training isn’t just increasing awareness. It’s changing behavior.
This is where the conversation shifts for CLOs. The challenge is no longer introducing AI into the organization. It’s building the capability to use it well, at scale and in ways that align with strategic business outcomes.
All that requires a different approach to learning.
Not one-off sessions or tool-based tutorials, but structured development that builds fluency over time. Not just for technical teams, but across leadership layers, especially those responsible for execution.
It also requires addressing a growing undercurrent of uncertainty.
As AI continues to reshape work, leaders are reckoning with what it means for their organizations and roles. A third (33 percent) have already eliminated or skipped opening a role in the past year because they believed AI could do the job, a figure that jumps to 52 percent in the technology sector. And the share of leaders who believe AI will replace most or all of their workforce within 10 years grew from 13 percent in 2025 to 20 percent in 2026.
Confidence in personal job security is also eroding at the leadership level, with only 56 percent of leaders saying they don’t believe AI will replace them within 10 years, down from 65 percent in 2024.
For CLOs, this undercurrent matters in more ways than one. Leaders navigating genuine uncertainty about their own relevance are harder to mobilize around transformation. Capability-building does more than just enable adoption in that environment. It gives leaders a concrete framework for engaging with AI — and a reason to do so.
Closing the gap with structured learning and development
Organizations that move from experimentation to enterprise-wide adoption won’t be the ones with the most advanced tools. They’ll be the ones that invest in capability systematically and at every level.
And that starts with leadership.
Building AI fluency among leaders creates a ripple effect of clearer direction, stronger use cases, more confident teams and, ultimately, more meaningful adoption.
This is the work that General Assembly and EZRA are focused on — helping organizations translate AI ambition into practical capability through structured learning and leadership development.
For CLOs, the opportunity is to lead that shift. To move beyond access and exposure toward fluency and application, and to ensure that the people responsible for driving transformation are equipped to do so.
Because the AI competency gap isn’t a technology problem. It’s a leadership one.
