Executive readiness
You can hire a Head of AI. Your culture will do to them exactly what it did to the software.
So should you hire one? Yes, and get clear on the job description first, because it depends on something most companies never check.
Should we hire a Head of AI?
A culture is a system, and that system is made of people who are each protecting something. Belonging. Competence. Security. The sense that they are good at their job. Relevance. Expertise. The way things have always been done.
So when somebody new walks in with a title and a mandate to change how the work happens, the organization does not push back because it is stubborn. It protects. And if the leadership team around that person has not evolved either, the new hire ends up where the tools did, attached to the outside of the same system.
Hiring the role is often how a company turns a leadership question into an org chart answer.
KPMG asked more than 2,000 senior leaders about this. Only 24% could name the CEO or executive committee as ultimately accountable for AI-informed decisions, and where that accountability is clear, established ROI runs 14% against 4%. Clarity beats headcount, and it costs less.
Find it, reframe it.
That is where adoption actually begins, and almost nobody starts there.
What skills will executives need as companies become AI-native?
Fewer technical ones than most people expect and more human ones than anybody planned for.
Think about how your executives got their seats. Deep expertise, years of industry judgment, a track record of good calls in a domain they know. All of that still matters. None of them were chosen because they were good at human change, at teaching new skills to people who did not ask for them, or at handling how the current way of doing things pushes back on a new tool when nobody has addressed what people are worried about.
Five things separate the executives and managers whose teams adopt from executives and managers whose teams comply.
- 01They orchestrate rather than directDesigning who and what produces an outcome, instead of holding the answer themselves.
- 02They make experimenting survivableA failed attempt becomes information the team keeps, rather than a story about the person who tried.
- 03They translateTurning what is happening into plain language, so nobody fills the silence with a worse story.
- 04They use the tools visiblyA team calibrates on what an executive or manager does far more than on what an executive or manager announces.
- 05They are the human in the loopEnough wisdom and enough confidence to back their own judgment against a confident-sounding answer that is wrong.
That fourth one is where leadership development and the cultural rewire turn out to be the same work. Culture is cumulative evidence. What an executive or manager praises, funds, interrupts, and is seen doing is the curriculum their people actually study.
The fifth is why industry expertise becomes more valuable rather than less. A general model treats collective patterns as truth and has no idea what is true in your industry or your building. It will be fluent and it will be wrong, and fluency is persuasive. The person who catches that is the one with twenty years in the work, which is precisely the person most worried about becoming redundant.
What does AI-native actually mean, and where is my company on the way there?
Five stages, and the word AI-native only describes the last one.
- 01 Individual productivity People use an assistant to write faster. Real value, entirely personal, and almost invisible on the P&L because nothing about how work moves has changed.
- 02 Your own knowledge, encoded A system built on your standards and judgment, so answers sound like your company rather than the average of the internet. The first stage that produces something a competitor cannot copy.
- 03 Into the workflow AI moves inside the work rather than beside it. Handoffs, drafts, invoicing, routing. It becomes visible in cycle time.
- 04 Redesign The work changes shape and roles get rebuilt around what humans are now needed for. Most companies skip this stage entirely, and it is also where people stop enjoying the help with mundane tasks and start realizing their job is genuinely changing. Where the ROI disappears
- 05 AI-native The organization operates differently because AI exists. Roles are defined, ownership of each tool’s outcomes is clear, and the culture no longer rejects the next tool because the fear has been addressed.
Almost no company sits on one rung. Sales may be at stage three while finance is at stage one, and the friction between them gets misdiagnosed as a people problem constantly.
Each rung asks something different of the executive or manager. Stage one needs permission. Stage three needs somebody who can redesign how work moves through a function. An executive or manager who was more than adequate at stage one can be genuinely underwater at stage three, and that is a capability gap rather than a performance problem.
Stage four is also where people stop loving the help with mundane tasks and start realizing their job is genuinely changing, with nobody internal equipped or free to handle what that brings up.
How ready is your leadership bench?
Probably less ready than it looks from the top, and the reason is more useful than the number.
Indeed and YouGov surveyed more than 1,000 job seekers and 300 hiring decision-makers between 22 May and 2 June 2026. 43% of managers said they felt poorly equipped or not equipped at all to lead AI-fluent employees. Among managers already leading AI-native talent, nearly nine in ten felt equipped.
Capability does not arrive through a bulletin or a session. It arrives through doing the work with support. That distance is not a gap in talent. It is a gap in reps, and reps can be designed.
There is a second gap worth knowing about. WalkMe’s State of Digital Adoption 2026, covering 3,750 leaders and workers, found 54% of workers had bypassed AI tools and done the task manually at least once in the past 30 days, with another 33% not using AI at all. Roughly eight in ten working around software their employer already paid for, and nobody sent a memo saying so.
Excitement is visible from the top. Changed workflows are not. The useful question is not whether your people are on board, it is how many of them have changed how they actually do their work.
Why do new ways of working get rejected?
Because a culture is not a mood or a set of values on a wall. It is a system that has been running a long time, and it is good at what it does.
That system has absorbed shocks, kept out what did not fit, and protected whatever the company decided years ago was worth protecting. When something lands that the existing logic has no slot for, the organization does what it has always done. It attaches the new thing to the outside and keeps running the way it was running.
A culture that pushes back on unvetted change is a culture doing its job. It has protected the company for years and it cannot tell the difference between a bad idea and a necessary one. It only knows what fits.
This is also why the situation does not respond to training. Cultural change is not a corporate process, it is a neural one. People move when the meaning is right, when the emotions around it are safe, and when their sense of who they are at work shifts to include the new thing. A module reaches none of those. It adds information to a system that was never information-limited, because your people already know what they are supposed to do. What they run into when they act on it is the part that has not changed.
Nobody gets replaced in any of this. The human evolves, and that is the entire job.
People do not sustainably sacrifice their values, and they will not describe the one they are defending to the person who signs their review.
So it gets surfaced from outside, reframed inside with the leadership team, worked one-on-one in each department, and then measured again a few months later to see whether it held. A new hire walking into an unchanged culture inherits all of this on day one.
So what is the actual decision?
Evolve the leadership layer. That is happening either way. The decision is how.
- Hire new executives and managers into the functions
- Slow, expensive at current talent prices, and it quietly tells the bench you have that they were found wanting. Most CEOs also discover the people who can do this are already doing it somewhere else.
- Hire someone to own it
- This is where the Head of AI conversation lands. That role without leadership development underneath gives you someone who owns the tools and has no authority over whether the people using them adopt or stall.
- Develop the bench you have
- Slower to feel and faster to hold, because those people already carry the industry knowledge, the relationships, and the trust. It means giving someone authority to change how decisions get made, and then actually taking something off their plate to make room. That last part rarely happens, which is why most internal versions quietly die in month three.
Either path is a real decision. What does not work is assigning it to someone who has no room for it and calling that a plan.
Whose job is this actually?
Ask whose job this is and the room goes quiet, and it is not because nobody cares.
HR is carrying compliance, hiring, benefits, difficult conversations, and whatever the last reorganization left behind. Learning and development is booked solid rolling out training for the tools already purchased. Your executives and managers are running functions and answering for numbers.
So the work does not get refused. It gets deferred, by capable people accurately assessing their own capacity, and deferral looks identical to unclear ROI or resistance from three levels up.
It does not sit cleanly inside any function either. HR owns policy and people processes rather than the operating logic of how decisions get made. The stack belongs to IT and throughput to operations. How executives and managers have traditionally worked belongs to everyone and to no one.
Where do you start?
Ask your leadership team one question at the next meeting and listen carefully to the room.
What have you personally shipped with AI in the last month, and who saw you do it?
The silence that follows, if it follows, tells you more than a readiness survey will.
Here is the limit of what an executive or manager can do alone. You can commit to modeling the behavior, and most capable executives will do it well. What you cannot do from your chair is see which of your executives and managers has quietly stopped proposing things, get an honest answer about fear from people who report to you, or judge your own authority architecture while standing inside it. Employees do not describe what they are protecting to the person who signs their review.
That gap is structural. It is the ordinary condition of standing inside a system while trying to describe it, and reading it accurately from outside is most of what a diagnostic is for.
How each executive and manager gets scored, and what happens next
Frequently asked questions
Should we hire a Head of AI?
Yes, once you are clear on what the job actually is. A Head of AI hired into an unchanged leadership layer ends up owning the tools with no authority over whether the people using them adopt or stall, which is exactly where the software ended up. KPMG found that where the CEO or executive committee is accountable for AI-informed decisions, 14% report established ROI against 4% where accountability is unclear. Clarity beats headcount, and it costs less.
What skills will executives need as companies become AI-native?
Fewer technical ones than most people expect and more human ones than anybody planned for. Running a team where part of the work is done by AI and part by people. Making experimentation survivable so a failed attempt becomes information rather than a story about the person who tried. Translating change into plain language at every level. Using the tools visibly. And being the human in the loop with enough confidence to say a fluent, confident-sounding answer is wrong.
What does AI-native actually mean?
AI-native describes the last of five stages. Individual productivity, then encoding the company’s own knowledge, then moving AI inside the workflow, then redesigning the work around what humans are now needed for, and finally an organization that operates differently because AI exists. Almost no company sits on one rung. Sales may be at stage three while finance is at stage one.
Why does our culture reject new ways of working?
Because a culture is a system that has been running on established logic for years, and that logic has protected the company by keeping out what did not fit. It cannot distinguish a bad idea from a necessary one, it only knows what fits. So when something arrives that the existing logic has no slot for, the organization attaches it to the outside and keeps running the way it was running.
If the company captures our expertise, does it still need us?
Yes. What gets captured is what someone already figured out. What cannot be captured is judgment about situations that have not happened yet, the ability to move people while they are frightened, and the authority to say the system got it wrong.
Start here
Before you write that job description
Who on your leadership team could describe what their own function looks like redesigned, rather than sped up? The Snapshot takes about five minutes. The call is with me directly, and there is nothing to prepare.
Not ready to talk yet
Get the next one in your inbox
People who find this page are usually reading something that finally names a piece nobody had named for them. Every week or two I publish one more of those — where AI adoption is actually breaking down inside companies, built from the research and from what I see in real leadership teams.
No pitch, no sequence designed to wear you down. If it stops being useful, one click and you are out.
Would you rather start with a read on your own organization? The AI Adoption Gap Snapshot™ takes five minutes and covers all twelve domains.