The first phase of AI was about tools.
The next phase is about being human.
For the past two years, most companies have focused on adoption: copilots, pilots, automation, productivity gains and AI roadmaps.
But the more important question is now becoming clear:
What kind of executive can actually lead a company in an AI-native world?
Because AI is not just changing workflows. It is changing what leadership means. And that is a far more uncomfortable conversation than which copilot licence to buy, which is exactly why so few boards are having it.
This week I want to define the answer: the AI-native CXO. What the job becomes, what it demands, and what today's leaders need to optimise for if they intend to still be in the room in five years.
How far are we from an organisation that runs almost entirely on AI?
Closer than most boards will admit, and not as close as the loudest voices on LinkedIn claim.
The fully autonomous company is still a story we tell, not a place we work. But the direction isn't in doubt. IBM's 2026 study of 2,000 global CEOs found that nearly a quarter of large organisations are already scaling agentic AI in at least one business function.
So the question isn't if. It's how fast, and who gets caught flat-footed.
And let's be honest about something most people won't say out loud: some big firms and some of their CISOs and CTOs are hiding behind "governance" and "risk" because they're scared. Scared that their own role is the one becoming obsolete. Slowing the company down feels like self-preservation. It isn't. It's a short stay of execution because sooner or later the board does the maths and works out that embedding AI across every corner of the business is the only way to keep the cost base competitive and keep pace with rivals who aren't hesitating.
Risk management is a leadership responsibility. It is not a hiding place.
The org chart is being rewritten from the bottom up
Most organisations already have an AI layer of some kind. What's changed is what sits underneath it.
In startups, the reflexive "let's hire three junior engineers" move is disappearing. US entry-level job postings fell around 35% between early 2023 and mid-2025, with AI-exposed roles like risk and compliance analysts among the hardest hit. PwC cut roughly 200 UK trainee roles because generative AI now does much of what first-years used to.
Cutting juniors to zero is still a mistake, and the sharp firms know it, juniors are usually the most fluent AI users in the building. But the shape of the work has changed for good. New joiners are being handed judgment calls on day one instead of three years of grunt work that used to build their instincts.
Which creates a problem,because I place senior people for a living: automate away every rung of the ladder and you stop manufacturing senior talent. The pipeline doesn't refill itself. Every leader cutting for this year's margin is also deciding who's available to hire in 2030.
And it fundamentally changes the leadership profile. If you're no longer managing a bench of juniors, you're managing AI, and that demands a different kind of leader. The new baseline is fluency. I suspect "AI-fluent" is about to become one of the biggest requirements in senior job specs over the next few years, and we're already watching it arrive a rung down: "fluent in using AI tools" has quietly become a core line in mid-level roles that didn't mention it eighteen months ago. It won't stay at mid-level for long.
It's not where you sit. It's whether you make the calls.
Middle management is compressed. Decision-making is decentralising; 79% of the CEOs in that IBM study said they're pushing accountability outward as AI spreads. The layers that existed only to move information up and down are thinning fast.
The C-suite, meanwhile, is the most insulated part of the org chart from direct automation. Strategic judgment, stakeholder trust, owning the hard call in the room that resists being codified into a workflow. We'll need CXOs for a good while yet, the job spec is changing.
Here's what's crucial to understand, and it's the whole point: it is not where you sit in the organisation that decides whether you make it through the AI era. It's how much of your role depends on judgment and the ability to make hard calls. A middle manager whose job is genuinely about judgment is safer than a CXO whose job was really about controlling information flow. At every level, the people who thrive are the ones the machine can't replace in the moment that actually counts: the hard call, made under uncertainty, owned afterwards.
Which is also why finding those people is about to become one of the most important jobs there is. In a world where judgment at the top is the scarcest asset a company has, hiring leadership talent for AI infrastructure and capital markets stops being a staffing function and the cost of getting it wrong, of backing a leader who merely looks the part, has never been higher.
That gap is where the AI-native CXO is born.
What the AI-native CXO actually is
1. They design a human-AI operating system but not runing a function. The old executive job was to own a department and optimise it. The new one is to design how humans and machines make decisions together across the whole business. AI-native leaders are architects of an operating model, not managers of a silo. Most very experienced people don't have this skill yet, and no one is going to teach it to them gently.
2. They decide at machine speed. AI compresses decision cycles brutally. When analysis that took two weeks takes twenty minutes, the bottleneck stops being information and becomes judgment; the willingness to decide, and to own it when it goes wrong. Executives who built careers on controlling information flow lose their edge overnight. The AI-native CXO's advantage isn't knowing things others don't. It's deciding faster than others dare to.
3. They work across dissolving boundaries. Over three-quarters of CEOs now say talent and technology leadership are converging. The CIO is becoming a Chief Intelligence Officer. The CFO is judged less on cost control and more on return on intelligence; the value the business actually extracts from its AI. The Chief AI Officer went from punchline to permanent fixture in barely a year.
4. They treat leverage as power, not headcount. For decades, team size was the measure of status. The AI-native CXO inverts it. The leader who delivers the same outcome with a quarter of the people and a well-orchestrated AI stack is now the one to promote. Empire-building has become a liability.
5. They are more human, not less. This is the part the automation panic misses. As machines take the mechanical work, the things they can't do become more valuable: judgment, taste, trust, the willingness to be accountable for an outcome you can't fully explain. The 83% of CEOs who said AI success depends more on people's adoption than on the technology are pointing at the same truth. The AI-native CXO isn't the most technical person in the room. They're the most human, sitting on top of the most capable machines.
What today's leaders should optimise for
If you want to still be leading and worth hiring in an AI-native company, optimise for these:
Judgment over knowledge. The answers are cheap now. Knowing which to trust, and when to overrule the machine, is not.
Speed of decision over control of information. Your old moat was knowing things others didn't. Your new one is deciding faster than others dare to.
Human connections over everything. In this world TRUST is the biggest currency you’d ever hold - so be bold and make those connections and start building trust.
Building the pipeline you'll need. Strip out every junior role for this year's margin and you'll pay a fortune importing senior talent in three. Design roles that still produce human judgment.
The irreducibly human. Trust, narrative, accountability, the ability to make people want to follow you into uncertainty. It's the part of the job with no substitute. Double down on it.
The executives who treat AI as an IT problem to be governed will be managed out by the ones who treat it as a Human Centric problem that needs to be solved.
The tools were phase one. Being Human is phase two.
Make sure you're the AI-native kind.
Sam - founder- Spearpoint search - formally workincrypto
See you next Friday.
…………………………………………………………………………………………………………………………………
If you enjoyed this, read more from our top posts:
