Tech-Fluent Leaders: What IBM's CEO Study Really Found

IBM’s 2026 Global CEO Study found that 85% of CEOs say all functional leaders must become technology experts in their domain, and 83% say AI success depends more on people’s adoption than on technology. Tech-fluent leaders, not better tools, are the new bottleneck.

Something I’m starting to observe in leadership conversations about AI: the statistics travel faster than their sources.

I caught myself doing it this week. I had picked up, from a YouTube video, that 76% of CEOs believe their leaders should be tech-fluent to apply AI. Good number. Quotable. I was ready to build an article around it. Then I did the thing we keep telling leaders to do with AI output: I went and checked the source.

The number I had was wrong. The real finding is stronger. And the study behind it says something more useful than the soundbite.

What the IBM CEO study actually found

The source is IBM’s 2026 Global CEO Study, run by the IBM Institute for Business Value with Oxford Economics. They surveyed 2,000 CEOs across 33 geographies and 21 industries between February and April 2026, making it one of the larger CEO datasets published this year.

Here is what the primary source actually reports:

• 85% of CEOs say all functional leaders must become technology experts in their domain. Not the CTO. Not the IT head. All functional leaders: sales, finance, HR, operations.
• 76% of organizations now have a Chief AI Officer, up from 26% in 2025. This is the number that gets swapped with the tech-fluency stat in some of the coverage online.
• 83% of CEOs say AI success depends more on people’s adoption than on the technology.
• CEOs estimate that only 25% of their workforce uses AI regularly, even though 86% believe their people have the skills to do it.
• Over half the workforce needs new skills. CEOs expect 29% of employees to need reskilling for different roles and 53% to need upskilling for their current roles between 2026 and 2028.

So the version I had heard undersells it: the demand is 85%, and it applies to every seat at the leadership table.

The mixed-up number is a small thing.

And that is still worth paying attention to, because the corrected picture changes the conclusion. CEOs are not saying they want tech-fluent leaders someday. They expect it now, they have already created executive roles for it, and adoption still is not moving.

Why tech-fluent leaders matter more than a Chief AI Officer

Look at those numbers side by side. 76% of these organizations hired a Chief AI Officer. The same CEOs estimate that only 25% of their employees actually use AI regularly. These companies bought the strategy role, the tools, and the training budgets, and the adoption still did not follow.

You have probably seen a version of this in your own company, right? A new system gets rolled out, the announcement is made, and three months later everyone is quietly back to the old way of working.

That is what makes the 85% finding the important one. A Chief AI Officer can set strategy, but adoption happens inside functions. It happens when the sales director changes how quotes get drafted, when the finance lead changes how reports get built, when the operations head changes how schedules get planned. If those leaders cannot personally judge what AI does well and where it fails, they cannot redesign their own workflows around it. They can only forward memos from the AI office.

Delegation is the quiet trap here. Technology used to be something a leader could hand to a department, the way payroll or facilities gets handed to a department. AI does not work that way, because the decisions it changes are the decisions inside each function: what gets automated, what gets reviewed, what gets promised to a customer. A leader who cannot evaluate the tool ends up approving things they cannot check.

The real question is whether the people who own each function can use AI, evaluate it, and govern it in their own domain. That is what the 85% are actually asking for.

What this means for Philippine businesses

For most Filipino SMEs, the Chief AI Officer conversation is theoretical. There is no budget line for that role, and there does not need to be. But that means the 85% finding lands directly on the owner and the leadership team. In a smaller company, you are the Chief AI Officer, whether the title exists or not.

I have seen this play out across the 100+ digital transformation projects I have led, and more recently in the AI workshops we run through the Philippine AI Business Association. The companies that move are not the ones with the biggest tools budget. They are the ones where the owner personally uses AI, where at least one department head has rebuilt a workflow around it, and where the leadership meeting includes questions like “what did we automate this month?”

We hold ourselves to the same standard at Tarkie and Otoma. Before we ask any customer to trust AI in their operations, our own leaders have to run their departments with it first. That practice has taught us more about adoption than any vendor pitch, because you feel the workflow problems personally instead of reading about them in a report.

The pattern from IBM’s data matches what we see on the ground here. Tools are available to everyone now. A one-person business in Batangas can use the same AI models as a multinational in Singapore. What separates them is leadership adoption speed. This is where smaller companies can play bigger: a 20-person company where the owner is fluent can out-adopt a 2,000-person company where fluency lives in one department.

What tech fluency actually looks like

Tech fluency does not mean learning to code. It does not mean reading model benchmarks. For a business leader, fluency shows up as a short list of practical abilities:

Use AI on your own work. Not through an assistant, not through a deck someone prepared. You, weekly, on tasks you actually do. Fluency comes from repetitions.

Judge output in your domain. A fluent finance leader can spot when an AI-generated analysis is confidently wrong. A fluent marketer can tell generic copy from copy that fits the brand. If you cannot evaluate the output, you cannot approve it responsibly.

Redesign one workflow at a time. Fluent leaders do not “add AI” to a process. They ask which steps the tool should absorb, which steps need human judgment, and rebuild the handoffs around that split. One redesigned workflow that the team actually uses teaches the organization more than ten pilots that quietly expire.

Ask vendors the uncomfortable questions. Where does our data go? What happens when the tool is wrong? What does this cost at ten times the volume? Fluency is also knowing what to ask before you sign.

Set the guardrails. Decide what data can and cannot go into public tools, who approves AI-assisted output, and how mistakes get caught. This is the part most leadership teams skip because it feels like bureaucracy. It is the opposite: clear guardrails are what make employees confident enough to use AI at all, because nobody wants to be the person who caused the incident. Adoption without governance creates the mistakes that set adoption back years.

None of this requires a technical background. All of it requires time on the tools, and that is exactly what most leadership calendars do not currently protect.

Where to start this week

If the IBM numbers describe your company, the starting move is small and personal.

Block one hour this week and run one recurring task through an AI tool yourself. Drafting a client proposal, summarizing a long report, preparing interview questions. Judge the result honestly. Then bring one question to your next leadership meeting: “Which of us can actually evaluate AI output in our own department?” The silence after that question is your real adoption gap, and it is a more honest metric than any tools inventory.

This is Level 1 work in The 4A AI Roadmap: leaders building personal fluency with AI assistants before the company attempts automation or agents. Skipping it is how companies end up in IBM’s statistics, with an AI office on the org chart and a 25% adoption estimate on the floor.

The IBM study settles the question of whether tech fluency is now part of the leadership job. An 85% consensus in a 2,000-CEO study is about as clear as survey data gets.

So the useful question is no longer: how capable is the AI?

It is: how fluent are the leaders who will run it?