A survey of 900 CEOs published last week made one thing unmistakably clear: the pressure has landed. Nearly three-quarters of U.S. CEOs are feeling it from their boards to prove AI-driven ROI. Eighty percent say their job is at risk if AI fails this year. Eighty-seven percent say their career is staked on AI’s success.

That pressure is legitimate. But it is pointed at the wrong question.

Most boards are asking: Is our AI working? The question with actual teeth is harder: Do you know what you are building, what it costs to get there, and what you will have when you arrive?

Because right now, most organizations cannot answer all three. They are investing in AI without defining the end state. Proving ROI on a destination they have not named. Measuring outputs without knowing which outputs signal progress toward what. Boards are holding CEOs accountable for a strategy whose success criteria have not been agreed on.

That is not an AI problem. It is a clarity problem. And it is more dangerous than most leadership teams realize.

Being AI-assisted, AI-enabled, and an agentic enterprise are not points on a single investment curve. They are three different destinations with three different price tags and three different definitions of success. AI-assisted is a productivity investment. AI-enabled is a process transformation investment. An agentic enterprise is an organizational redesign investment. The capital required is different. The timeline is different. The capability you build is different. And critically, the outcome you can claim is different.

A company optimizing for the first will not accidentally arrive at the third. Conflating them does not split the difference. It produces organizations that are over-spending on tools, under-investing in architecture, and reporting efficiency gains to boards that were quietly expecting transformation.

The Hidden Cost Nobody Is Reporting

Start where the majority of organizations are: AI-assisted work. Copilot. Claude. Gemini embedded in the tools people already use. Individual augmentation at scale.

The efficiency case is real. But there is a gap between efficiency captured and efficiency realized that almost no one is measuring. When a knowledge worker saves two hours through AI assistance, those two hours do not automatically become two hours of higher-value work. The work that consumed them has to be consciously redesigned, redirected, and replaced. In most organizations, it isn’t. The time gains leak into the existing backlog. The ROI that boards are demanding is sitting in a gap between what AI made possible and what the organization chose to redesign.

There is a second cost, less visible and more structural. The half-life of technical skills is shrinking. New models, new capabilities, new tool releases arrive at least monthly. Keeping up is no longer optional. It is a job requirement that has not yet been written into any job description. A realistic estimate for knowledge workers: at least ten percent of working time each week now needs to go toward learning and applying what was last released. Not aspirational. As a design constraint.

That is not a training problem. It is a work design problem. It belongs in the workforce strategy, not the L&D budget. Organizations that treat it as a training expense will fall behind organizations that treat it as an operating reality.

AI-assisted is not a destination. It is a starting condition with a ceiling: roughly ten to thirty percent productivity gains in processes that were not redesigned to capture them. That is worth having. It is not worth betting a CEO’s career on.

Three Types of AI Organizations

Here is the distinction that most AI conversations collapse: having AI in your organization and choosing what kind of AI organization to become are not the same decision. There are three meaningfully different outcomes, and they require different investments, different timelines, and different definitions of success.

AI-Assisted. Humans do the work. AI helps them do it faster, with better inputs, with less friction on routine tasks. Value is real. The ceiling is real. Nothing about the organization’s structure, economics, or operating model needs to change to get here. Most organizations are somewhere in this territory right now.

AI-Enabled. Agents run in production. Processes are automated. Parts of the business operate differently because AI is doing meaningful work end-to-end in specific workflows. Gartner projects that forty percent of enterprise applications will include task-specific AI agents by the end of 2026, up from less than five percent in 2025. The infrastructure ecosystem is maturing fast. New enterprise services firms backed by major private equity are forming specifically to bring agents into the core operations of mid-sized companies. This is legitimate and valuable work. It is also bounded. Agents in production workflows, without the organizational architecture to run on them, is an enhanced company. Not an agentic one. The gains are real. The ceiling is still there.

Agentic Enterprise. This is a different kind of company. Not a company that has deployed agents. A company that has been rebuilt around them. The organizational logic has changed: how work is structured, how teams are designed, how economics are governed, how leadership makes decisions. All of it now reflects a world where agents are participants in every significant business process. This is not the natural endpoint of deploying more agents. You cannot arrive here by accident. It requires a deliberate architectural choice, made at the top of the organization, before the infrastructure is built.

The decision most organizations have not made is which of these three they are pursuing. Boards are measuring AI performance. Most of them have not defined which outcome they contracted for.

What an Agentic Enterprise Actually Is

An agentic enterprise is not defined by how many agents it has deployed. It is defined by whether the organization has been redesigned to run on them.

Six things have to be true simultaneously.

Future-focused leadership. The leadership team has defined not what agents do for the company, but who the company becomes with agents in it. Strategy, capital allocation, and new adaptive leadership behaviors and skills have all changed to reflect that answer.

Intelligence-led value. The business model has been redesigned for a world where agents are a channel, a product, and a value-creation engine. The question is not where AI reduces cost. It is where AI creates revenue that did not exist before.

Human-AI teams as the unit of work. Teams are structured around operating, supervising, and orchestrating agents rather than executing work manually. People are not users of AI tools. They are accountable partners in a workforce that includes agents.

Adaptive business architecture. The organization can re-compose its capabilities in weeks rather than years. Not through reorganization. Through modular redesign that treats structural change as a continuous competency, not a periodic event.

Agentic FinOps. Agent economics are governed at the task level, with clear cost attribution per workflow, per agent, per outcome. If you cannot answer what a single agentic task costs end-to-end, the architecture is not governed.

Agentic platform and systems. The technical foundation was built for agents from the ground up, not retrofitted to accommodate them. Data, integration, observability, and security are designed as agent infrastructure, not adapted from what existed before.

Every one of these is a leadership decision before it is a technology decision. Gartner’s most recent analysis forecasts that more than forty percent of agentic AI projects will be canceled by the end of 2027, due to escalating costs, unclear business value, or inadequate risk controls. That is not a technology failure rate. It is a strategy failure rate. Organizations are building agent infrastructure without first deciding what kind of organization they are building toward.

The Workforce Question Most Strategies Miss Entirely

The technology conversation and the workforce conversation are being held in separate rooms. In most organizations, the people strategy is trailing the AI roadmap by six to twelve months. That gap is where value erodes, where trust breaks down, and where the talent pipeline gets quietly damaged in ways that take years to surface.

The framing that clarifies this most sharply: Human by Choice.

As agents scale, the instinct is to ask where humans are no longer needed. That is the wrong question, and it produces the wrong answers. The right question is: where do humans stay in the value chain by design, because their presence is a source of competitive differentiation or because their absence creates unacceptable risk?

Those are two very different reasons to keep a human in a process. Both are legitimate. Neither is residual.

Trust sensitivity. Consequence irreversibility. Brand exposure. Judgment complexity. These are not soft considerations. They are the criteria that determine whether a visible human makes the result more credible, whether an AI error would be reputationally catastrophic, whether a mistake can be undone. They are the inputs to a deliberate placement decision, not default positions held over from the previous operating model.

Where humans stay is strategic, not accidental.

The organizations that skip this question are not avoiding a strategic debate. They are making three expensive decisions by omission.

First, AI intensifies work; it does not reduce it. Recent HBR research confirms what practitioners are already experiencing: AI tools consistently produce more work, not less. Employees work faster, take on broader tasks, and extend into more hours, often without being asked. The original task gets done at sixty to sixty-five percent of the previous time. The remaining capacity does not go to rest. It goes to new insights, new analysis, new surface area the team never had time to explore before. Total time investment goes up. This is not a complaint. It is a design constraint. Workforce strategies that assume efficiency gains translate directly to capacity reduction are building on a false premise.

Second, agents absorb entry-level work, and entry-level work is the apprenticeship path. When agents handle the tasks that junior employees used to do, the development pipeline that produced senior judgment breaks. The next generation of practitioners never gets the repetitions that build pattern recognition, error detection, and domain expertise. The organization will feel this three years from now, when the people who were supposed to develop into senior roles did not, because the work that would have developed them was automated before anyone thought through the implications. Intentional learning design, including reserving specific tasks for human practice and building deliberate feedback loops, is a workforce infrastructure decision, not an HR afterthought.

Third, accountability does not transfer to agents by default. It diffuses, which is worse. When a human approves ninety-five percent of agent recommendations without meaningful interrogation, the organization has created the appearance of oversight without its substance. The accountability sink: nominally, someone is responsible. Functionally, no one is exercising judgment in the way the system assumed. This is how “work-slop” proliferates, how brand-inconsistent outputs make it to market, and how errors that should have been caught compound invisibly until they become visible in the wrong context.

The workforce architecture of an agentic enterprise is not about headcount. It is about decision architecture: who predicts, who judges, who acts, and who owns outcomes. In every significant process, those four questions need a clear answer. Where they do not, the organization has an accountability sink, not an agentic capability.

What the Board Conversation Should Actually Cover

Most board AI discussions focus on investment, risk, and competitive positioning at the macro level. The questions that reveal whether an organization knows what it is building are more specific.

Can every executive on the leadership team state in one sentence who the company becomes with agents in it? Not what agents do for efficiency. Who the company becomes. This is the strategic intent question, and most leadership teams cannot answer it consistently. That inconsistency is the thing boards should be concerned about.

Has capital been reallocated in the last two quarters to match that answer? Not approved in principle. Actually moved. Capital allocation is the most reliable signal of whether an AI strategy is real or aspirational.

Are workflows being designed for agents, or are agents being accommodated into existing processes? The difference between retrofitting and redesigning from the ground up is the difference between AI-Enabled and Agentic on the maturity curve. One has a thirty percent ceiling. The other does not.

Do you know what a single agentic task costs end-to-end, including compute, memory, and tool calls? If not, you have agents. You do not have a governed agentic capability.

For the people strategy specifically: does your workforce design include a view of where humans stay by design, with clear reasoning for each placement decision? Or does human involvement default to wherever agents have not yet been deployed?

The board also needs a shared answer to one more question, and most do not have it: what should we expect, and when?

The answer depends entirely on which destination the organization has chosen. AI-assisted returns show up in months, not years. Productivity gains, reduction in manual effort, faster output cycles. These are measurable quickly, but they are also bounded. If the board is expecting them by Q3, that is reasonable. If the board is expecting them to compound into competitive advantage, that is a different investment they have not yet made.

AI-enabled returns take longer to materialize and require more organizational change to capture. Process automation at the department or multi-department level, redesigned workflows, new value chain capabilities. The reasonable horizon is twelve to twenty-four months from committed investment, not from the pilot. The mistake most organizations make here is measuring ROI from the proof-of-concept rather than from the point at which the work was redesigned around the new capability.

An agentic enterprise does not produce a single ROI moment. It produces a compounding architecture. The early returns look modest relative to the investment. The gains accelerate as the system matures, as agent economics improve, as human-agent team fluency increases, and as the organization’s capacity to re-compose around new opportunities shortens. The board needs to understand that the ROI curve for an agentic enterprise is not linear and is not front-loaded. Organizations that evaluate it on a twelve-month payback horizon will exit the investment before it pays.

If the board has not agreed on which curve it is on, it cannot set appropriate expectations, allocate capital correctly, or evaluate progress honestly. That misalignment is not a reporting problem. It is a governance problem.

Where to Start

The organizations that will reach the agentic enterprise are not the ones that move fastest on technology. They are the ones that answer the strategic intent question first and build the architecture to match.

That sequence matters. Agentic platforms and systems built without clarity on human-agent workforce design will create governance debt that compounds. Workforce redesigns built without clarity on the operating model will produce role changes without infrastructure to support them. The six components of an agentic enterprise are interdependent. None of them works well in isolation.

The prerequisite is leadership alignment on three things: where human judgment creates competitive advantage and should be protected, what the organization’s boundaries are for agent autonomy, and what the talent mix looks like across fully human, human-agent, and fully autonomous work. These are not technology decisions. They are the decisions that make every technology decision coherent.

Every organization is building something with AI right now. Most of them have not decided what.

The question that determines which category you are in five years is not how much you are spending on AI, or how many agents you have deployed, or what your AI ROI looked like in Q2. It is whether you know, right now, what you are building toward.

What answer does your leadership team give when you ask which type of AI organization you are building? And more importantly: is the answer consistent across the room?

U.S. organizations spent $102.8 billion on training in 2025. They allocated 4% of that to AI skills. Meanwhile, organizations are pouring $85,000 a month or more into AI tools and planning to increase their AI investment by 92% over the next three years.

The money is going to the machines. Not the humans who need to work with them.

This isn’t a training gap. It’s a fundamental misunderstanding of what’s required to realize the promise of AI to the organization. And the cost of getting it wrong is not a bad quarter. It’s organizational irrelevance.

The numbers don’t lie. But they do mislead.

Here’s what the current landscape looks like. Organizations are spending an average of $874 per learner annually on all training. Training hours per employee actually dropped from 47 to 40 in a single year. Only 27% of firms have a structured AI upskilling program. And 44% of companies reskilled less than 5% of their workforce over the past year.

At the same time, enterprise AI budgets are growing 75% year over year. The average organization is now spending over a million dollars annually on AI tools alone.

One side of the ledger is accelerating. The other is barely moving.

But here’s what makes this worse. Most of what organizations call “AI training” is tool training. How to use the chatbot. How to write a prompt. How to navigate the new interface. That is not reskilling. That is not redesigning work. That is not going to lead to ROI on the AI spend. That is onboarding to software. And it is wildly insufficient for the AI available transforming businesses today.

Training teaches you how to use a tool. Reskilling changes how you think about your work.

The World Economic Forum estimates that 44% of workers’ skills will be disrupted in the next five years. BCG’s research found that half of frontline employees have hit a “silicon ceiling,” stuck at basic AI usage with no pathway to deeper integration.

The distinction between training and reskilling matters because AI agents are approaching PhD-level competence in functional domains. Engineering. Finance. Supply chain. Marketing analytics. Legal research. The work that most knowledge workers spent decades learning to do is becoming table stakes for a well-orchestrated set of agents. The question is no longer “can you do this work?” It’s “can you evaluate, direct, and improve upon what AI produces?”

That requires a different set of skills entirely.

Three types of skills that matter now

1. Human skills. Stop calling them soft.

These are about to become the most valuable skills in any organization. They are also the hardest to develop. Here are the ten that matter most.

Collaboration and facilitation. Aligning humans around decisions that include AI. Managing conflict, creating shared understanding, and enabling teams to adopt new ways of working. The technology is the easy part. Getting people to change how they work together is the real challenge.

Vision and inspiration. Setting direction, motivating change, and using AI to accelerate a compelling narrative, not just efficiency. The leaders who win in this era will not be the ones who automate the most. They will be the ones who inspire their organizations to become something new.

Systems thinking. Understanding second-order effects, dependencies, incentives, and how AI changes a whole process, not just one task. This is the skill that prevents local optimization from creating organizational chaos.

Judgment under uncertainty. Making sound calls with incomplete information. Balancing risk, upside, ethics, and context when the data doesn’t give you a clean answer. AI can generate options. It cannot weigh them against the political, cultural, and strategic realities of your organization.

Problem framing. Defining the real question, constraints, success criteria, and what “good” looks like before you ever prompt. The quality of what AI produces is a direct function of how well the human frames the problem. Most people skip this entirely.

Critical thinking and verification. Interrogating outputs, spotting gaps, testing assumptions, triangulating sources, and knowing when not to trust the model. This is the skill that separates someone who uses AI from someone who is used by it.

Domain discernment. Knowing what matters in your domain: the edge cases, the regulations, the customer nuance, the things that don’t show up in training data. Translating “generic” AI output into “situationally correct” is where deep expertise meets practical value.

Clear communication. Expressing intent and outcomes crisply. Writing, speaking, and translating complexity for different audiences, both humans and machines. In a world of AI-generated content, the ability to communicate with precision and clarity becomes a differentiator, not a baseline.

Prompting and interaction design. Structuring inputs, iterating, giving feedback, and designing workflows where AI is a teammate, not a vending machine. This is the bridge between human skills and technical skills. It is where most people are today. It is not where they need to stay.

Taste and quality standards. Recognizing what’s excellent versus merely plausible. Aesthetic sense, craft, and discernment about coherence, voice, and fit. AI produces “good enough” at scale. Humans define what “great” looks like.

Here’s the truth. When your AI agent can produce PhD-level financial analysis in seconds, the person who wins is the one who knows what to do with that analysis. Who can see what’s missing. Who can connect it to a strategy that hasn’t been articulated yet. Who can walk into a room and make a decision when the data is ambiguous.

2. Technical skills. Beyond the chat window.

Most organizations think “AI technical skills” means learning to prompt. That was 2023. The skill now is learning to build new workflows. Managing agent orchestration. Understanding how to chain tools together to produce outcomes that are 5 to 10 times faster than legacy processes.

More importantly, the real technical skill is learning how to learn. The tools change monthly. Features ship weekly. A 12-week bootcamp is obsolete by week six. The organizations that win here are not the ones with the best curriculum. They are the ones whose people can pick up a new capability on Monday and have it integrated into their workflow by Friday.

This doesn’t come from training. It comes from team-based experimentation and collaborative discovery.

3. Functional skills. The ones your agents are about to master.

Engineering. Finance. HR. Supply chain. Marketing. Product strategy. These are the domains where most of us have built our careers, going deep over years and decades. Where we have created specialization of roles.

AI agents are rapidly closing that gap. Not replacing the need for domain knowledge, but fundamentally changing the nature of what it means to be skilled in these areas. The skill is no longer “I can build this financial model.” It’s “I have the skills to evaluate the model my agent produced, challenge its assumptions, and know when it’s wrong.”

This is where functional expertise evolves from execution to orchestration. The knowledge still matters. But how you apply it changes entirely.

How you actually redesign your work (hint: it is not a training program)

Let’s be honest about something. Most knowledge workers using AI chatbots are already finding ~8 hours of reclaimed time per week. They’re getting drafts faster, research faster, and analysis faster.

Where is that time going?

In most organizations, nowhere visible. It becomes productivity leakage. It doesn’t show up on a balance sheet. It can’t be redeployed because leadership doesn’t know it exists. People are getting faster without the organization capturing any of that value.

This is the starting point for reskilling to redesign work. That found time isn’t a bonus. It’s the raw material for transformation. The new expectation for knowledge work should be clear: one hour a day of experimentation. Hacking your own job with AI. Not as a side project. As a core part of how you work.

The goal is not incremental improvement. It’s 5 to 10x shifts in how work gets done in your function. New workflows. New ways of collaborating. New outputs that weren’t possible six months ago. But this only happens when people have access to the right tooling, understand how to build agentic workflows, manage agents, and work collaboratively as roles evolve beyond traditional boundaries.

Here’s what this looks like in practice.

Monday: Focus and align. The team picks a specific area of their work to “hack” that week. Not a vague goal. A specific process, deliverable, or workflow. Alignment as a group on what they’re trying to improve and why. Embedded AI experts hold office hours to help people get unstuck and push past the chat window into real workflow design.

Tuesday through Thursday: Experiment. People try new approaches. They build agentic workflows. They test. They fail. They iterate. This isn’t theoretical. It’s working on real work with real stakes, just with new methods. The embedded expert is available, not running a class, but coaching in context.

Friday: Show and share. The team comes together. What worked? What didn’t? What produced a genuine breakthrough? The group agrees on the best solutions to scale across the team. The winning experiment becomes next week’s standard operating procedure. Then the cycle starts again.

This becomes a new way of working. Not a training initiative. A working rhythm. A permanent operating cadence where improvement is built into the week, not bolted on as a quarterly workshop.

Skills are the new currency. Not roles. Not titles.

As this experimentation takes hold, something uncomfortable happens. Roles start to blur. The marketing analyst who built an agentic workflow for competitive intelligence is now doing work that used to belong to the strategy team. The finance associate who automated reporting is now spending time on analysis that was previously reserved for senior leaders.

This is where we need HR to support architecting a skills-based organization.

Skills are the new currency in which organizations will trade. Not roles. Not titles. Not org chart boxes.

This means building a skills-based architecture that can identify which skills are needed, which skills exist across team members, develop those skills through embedded practice not just courses, test for proficiency in meaningful ways, visualize the skills landscape across the organization and deploy team members on the most important work to the organization.

When you can see skills clearly, you unlock something powerful: an internal skills marketplace. Talent stops being hidden inside departments. The person in operations who taught themself agent orchestration becomes visible to the product team that needs exactly that capability. Work gets matched to skills, not to job descriptions written three years ago for a world that no longer exists.

The wall you’ll hit. And what comes next.

Here’s the truth that most leaders don’t want to hear. Everything I’ve described above is relatively achievable at the department level. A progressive leader with access to the right tools, data and the conviction can transform how their team works.

The hard part is what comes next.

When experimentation works, roles start collapsing across departments, not just within them. The value chain itself starts to compress. And that’s where organizational politics becomes the real blocker. Not the technology. Not the data. Not the skills gap. The turf battles.

Who owns the workflow that spans marketing and sales? Who decides when the finance team’s agentic process replaces half of what procurement does manually? When roles blur across departments, who decides?

This is where the real value lives that boards and the street is looking for. The reimagined value chain. The 10x efficiency gains. The new business models that create new revenue not just do what you do today faster. But it’s also where most organizations will stall, trapped by inertia and internal politics.

The $102 billion question

The question isn’t how much you’re spending on training. It’s whether you’re developing the skills that matter when AI can do the work your people spent decades learning to do. And whether you’re brave enough to redesign work itself, not just teach people new tools.

The organizations that figure this out won’t just adapt. They’ll be the ones re-founding their organizations.

What’s working in your organization? Are you seeing the shift from training programs to genuine reskilling and work redesign?

Subscribe now

Share

Enterprise AI budgets are growing with the average organization now spending over a million dollars annually on AI tools alone. Customer satisfaction just hit a four-year low. Consumer trust in AI peaked in 2023 and has been declining since. And most companies are measuring success by how many customer interactions never reach a human.

That is not an efficiency strategy. That is a brand erosion strategy.

Here is what I think most leaders are getting wrong about the human side of AI. They are treating “where humans show up” as a cost optimization question. It is not. It is a brand architecture decision, a risk management strategy, and the thing that will determine whether your organization can reinvent itself.

Let AI do what AI does well. That is not the argument.

Let me be clear about something. This is not a case against automation. AI agents should absolutely handle routine transactions, data processing, report generation, scheduling, and the thousand small tasks that consume knowledge workers’ days. That work should get faster, cheaper, and more consistent. Organizations that fail to automate the automatable will fall behind. Full stop.

The argument is about what happens next. When AI absorbs the routine, the interactions that remain human become disproportionately important. They are the escalation. The ambiguity. The high-stakes negotiation. The moment a customer decides whether they trust you. The decision that carries regulatory exposure. The conversation that determines whether a team pivots or stalls.

These are not leftover interactions. They are the ones that define your brand, your risk posture, and your capacity to change. And most organizations are staffing them with people whose judgment, communication, and critical thinking skills have been quietly degrading because nobody designed their work to keep those capabilities sharp.

The moments that matter are now the only moments.

Avaya’s 2026 research found that 90% of consumers believe they should have access to a real person when needed. Eighty percent prefer humans for trust-driven interactions like dispute resolution. And 73% would leave a brand entirely if it offers only AI interactions with no human alternative.

Meanwhile, Forrester’s 2025 CX Index shows satisfaction scores at a new low for the fourth consecutive year. Fast Company’s analysis put the diagnosis plainly: organizations fell into a “containment trap,” measuring success by how many interactions they kept away from a human. On paper, it looked efficient. In practice, it destroyed trust.

Gartner’s data tells the other side of the story. Buyers who feel an experience has been genuinely personalized are 1.8 times more likely to pay a premium and 3.7 times more likely to buy more than planned. That personalization is not just an algorithm recommending the right product. It is the human who knows your account, catches the nuance, and makes the call that a model cannot.

“Human by Choice” means identifying which moments carry that weight, staffing them with your best people, and investing in the capabilities that make those interactions exceptional. Not as an afterthought to your AI strategy. As the centerpiece of your brand.

Human oversight is now a fiduciary obligation.

The regulatory landscape changed faster than most boards realize. The EU AI Act, effective August 2026, classifies human oversight as mandatory for high-risk AI systems. FINRA has signaled it will examine whether human oversight is substantive or merely nominal, with compliance frameworks expected by Q4 2026. Aon reports that more than 90% of insurance decision-makers now consider AI-driven incidents a material concern.

The 2026 International AI Safety Report, authored by over 100 experts across 30 countries, made the point that should keep executives up at night. IBM’s Francesca Rossi, who contributed to the report, stated that a nominal “human-in-the-loop” approach is not enough, warning that if humans are overloaded or lack the right information, oversight becomes symbolic.

Symbolic oversight with real liability exposure. That is where most organizations are headed.

If your people cannot exercise independent judgment on AI outputs, because their critical thinking has atrophied or because they were never developed in the first place, your “human in the loop” is a fiction. Regulators are not going to accept that fiction. Courts are not going to accept it. And your insurance underwriter is already pricing that risk into your coverage.

“Human by Choice” in this context means maintaining the organizational capacity for genuine human judgment. Not as a speed bump in the automation pipeline. As your last line of defense against the errors, biases, and cascading failures that agentic AI systems will inevitably produce.

You cannot preserve what you do not exercise.

This is where the brand argument and the risk argument converge into something more urgent.

Gartner’s strategic predictions warn that 50% of organizations will require “AI-free” skills assessments by 2026 because critical thinking is degrading from overreliance on generative AI. The IBM Safety Report researchers flagged something even more troubling: the greater danger is not a dramatic loss of control but the slow normalization of dependency, where organizations cede autonomy voluntarily, one workflow at a time.

IDC projects that over 90% of global enterprises will face critical skills shortages by 2026. Yet only 35% of leaders feel they have prepared employees effectively. The investment is going to the machines. Not the humans who need to work alongside them.

This is the atrophy trap. When you design every workflow to minimize human involvement, judgment degrades. When judgment degrades, your brand moments get worse. Your risk exposure grows. And your capacity for reinvention disappears. You end up efficient, fast, and completely unable to change direction.

The skills that drive reinvention, empathy, creative problem-solving, ethical reasoning, stakeholder negotiation, systems thinking, are not things you can buy back with a training program after you have let them erode for three years. They are built through practice, friction, and stakes. They compound with investment. They atrophy without it.

Three decisions most executives are avoiding.

First: Where will you make “human by choice” a brand promise?

Not everywhere. Deliberately. Audit your customer journey and identify the three to five moments with the highest emotional or financial stakes. The escalation. The renewal. The moment of crisis. Designate those as human-committed interactions. Set quality standards for them. Staff them with your best people and measure them as brand investments, not cost centers. Then make that commitment visible to your customers. In a market where AI interactions are becoming the default, “you will talk to a person who knows your situation” is a brand position with real pricing power.

Second: Where is human oversight a genuine control, not theater?

Map your AI deployment to your risk exposure. Classify every AI-assisted decision by two criteria: reversibility and consequence. Irreversible, high-consequence decisions need substantive human oversight from people with real domain expertise. That means funding their ongoing development, testing their ability to catch AI errors (not just rubber-stamp outputs), and building escalation paths that are fast enough to matter. If your human-in-the-loop cannot explain why they approved or rejected an AI recommendation, your oversight is decorative.

Third: Who owns human capability as a strategic asset?

This is not an training question. It is a C-suite design question. Assign explicit executive accountability for the long-term health of human capability in your organization. Build a human capability dashboard that sits alongside your AI adoption dashboard. Track whether judgment, critical thinking, and ethical reasoning are being practiced or just presumed. Fund development at a ratio that reflects the strategic importance of these skills. Because the board will eventually ask who was responsible when the human oversight failed. And “we assumed our people could still do that” is not an answer that will hold up.

The bottom line.

The organizations that deploy the most AI will not be the ones that win. The ones that win will be the ones that were deliberate about where they stayed human, invested deeply in making those human moments exceptional, and protected the capacity for judgment that AI cannot provide and regulation now requires.

“Human by Choice” is not a category on a skills slide. It is a strategic position. And right now, most companies are choosing it by accident, if they are choosing it at all.

Where in your organization is the human interaction a premium brand moment? And is anyone investing in it that way?

Subscribe now

Share

Here is what happened at Chegg.

For years, the company dominated the homework-help market. It had the brand, the users, the revenue model, and the data. When AI arrived, Chegg’s leadership did not ignore it. They moved fast. The CEO met with Sam Altman personally, partnered with OpenAI, and launched CheggMate, an AI-powered learning assistant. In an SEC filing he wrote: “We are embracing it aggressively and immediately.”

They ran the AI play. They optimized.

Then their customers started using AI instead of Chegg. Not a competitor. Not a new entrant with a better product. The technology itself became the substitute. Revenue dropped 36% year over year. The subscriber base fell 40%. By late 2025, Chegg had cut 56% of its global workforce. Not because it failed to implement AI. Because it never asked what kind of company it needed to become when AI changed what its customers needed.

Chegg is not a cautionary tale about AI adoption being slow.

It is a cautionary tale about leaders asking the wrong question. They asked: how do we use AI to improve what we do?

They should have asked: given what AI now makes possible, does what we do still matter? Where will value be created? How will we participate? What are we willing to let go of?

Those are not the same questions. And right now, almost every board and CEO I talk to is answering the first one while the second one goes unasked.

Doing what you have always done faster and cheaper is only a winning strategy if your market, your customers, and your competitive landscape are staying still. They are not. They may not be changing as fast as Chegg but every market is changing.

This is not a technology implementation problem. It is a re-founding imperative.

The disruption isn’t inside your org chart

Kodak invented the digital camera. Blockbuster had multiple chances to acquire Netflix. These are not stories of companies that lacked resources or intelligence. They are stories of leaders who optimized brilliantly for the world they understood while a different world was being built around them.

The real risk of AI is not that your operating costs stay too high. The risk is that you spend two years running an internal efficiency play while a competitor who didn’t exist three years ago is redesigning the value proposition your customers actually want. The risk is the adjacent business you dismissed because “that isn’t what we do.” The risk is waking up in three years having successfully optimized a platform your market no longer needs.

That is the conversation almost no board is having. Instead, the CTO owns the AI roadmap. The risk committee reviews the AI register. The CEO gets briefed on deployment milestones. And nobody asks: what kind of company do we need to become?

So what is re-founding, exactly?

The concept has a lineage worth understanding. Reid Hoffman first argued in 2012 that “founder” is a state of mind, not a title. By 2024, on Masters of Scale, he had extended that logic into a formal definition: a re-founder joins an organization as a catalyst for evolving culture, roadmap, and strategy while preserving the valuable building blocks of the business. Critically, re-founders take the kinds of risks that matter to the company that will exist after their tenure, not just to the current quarter.

Re-founding is not transformation. Transformation keeps the destination in the same zip code. You improve the operating model, digitize the processes, upgrade the talent. The company looks similar. It just runs better.

Re-founding asks something more uncomfortable. It asks what you are willing to let go of. It asks what the company needs to become, not what it needs to improve. It requires naming what must die: the incentives, the assumptions, the role definitions, the business line, the success metrics that were built for a world that is being replaced. Hoffman’s warning is precise: the mistake is either preserving everything or treating the company as a blank page. The discipline is knowing the difference.

And in the AI era, that discipline now belongs to every incumbent CEO. Satya Nadella is the strongest operating example. When he became Microsoft’s third CEO, he knew the job could not mean doing what Gates or Ballmer had done. He reset purpose, mission, and culture first. Then he let the platform shift rebuild everything else around it. In Microsoft’s own words, the task was simultaneous unlearning, learning, and mission reimagination for the AI era. Decades of change compressed into a few years.

Microsoft’s current AI posture is not the result of a great AI strategy. It is the result of a leader who understood that re-founding had to come before road-mapping.

That sequence matters. And most organizations have it completely backwards.

AI doesn’t just change how work gets done. It changes five things simultaneously: what customers value, how fast business models can shift, how work is performed, what infrastructure becomes strategic, and what governance burden now sits directly with the CEO and board. That is not a technology problem. That is a re-founding condition. And it requires a re-founding response.

The five-move re-founding playbook

Move 1: Define your Horizon 3 identity before your Horizon 1 roadmap.

Run a structured session with the CEO and board together, not separately. The question on the table: Where is the world going? Where will value be added? What will we let go of? What business do we actually want to be building?

Name the business lines you have avoided because “that isn’t what we do.” Test whether that logic still holds. Define the future identity specifically enough to create tension with today’s org chart, operating model, and investment priorities. If it doesn’t create tension, it isn’t a re-founding identity. It is a rebrand.

Move 2: Name the moment, not the initiative.

The CEO makes a public, unambiguous declaration: this is a re-founding condition, not an upgrade cycle. Not a vision statement. A strategic condition. It goes in the all-hands, the board letter, and the annual report.

The board formally expands its mandate from AI oversight to re-founding oversight. Separate agenda time. Separate risk framing. A longer time horizon than the next two quarters.

It is more than communication and management: capital allocation shifts visibly. If the AI budget still sits inside IT or the innovation lab, the organization will read that signal correctly: this is still a technology initiative.

Move 3: Decide what you will preserve, then rewrite how it shows up.

Re-founding is not a blank page. A few core values and focus areas survive. But here is what most organizations miss: their behavioral expression has to be rebuilt for an AI-enabled organization.

Values printed on a wall before 2023, without AI-era behaviors attached to them, are not a culture anchor. They are a source of organizational incongruence. What does “accountability” mean when an AI system made the first three decisions in a process? What does “customer obsession” mean when your front line is a human-AI team? Keeping outdated behavioral norms while adding new ones doesn’t build a bridge. It creates confusion about what is expected. And your people will feel that confusion before your leadership team does.

Move 4: Make unlearning explicit, starting with yourself.

Here is the executive mirror test. If any C-suite leader says “that’s for my teams, or I don’t see much use for AI in my own work,” you are not re-founding. You are implementing technology, and not well.

Every executive, including the CEO, identifies at least two specific ways their own decision-making and daily work practice changes with AI. This is not a delegation exercise. It is a personal reskilling commitment made visible to the organization.

Build unlearning into leadership rituals: what assumptions, processes, and success metrics are we formally retiring this quarter?

Fund reskilling at every level with the same seriousness as capital investment. If the reskilling budget is a rounding error next to the AI tooling budget, your organization will fill that gap with anxiety, resistance, and the quiet exit of your best people.

Every job changes. Every leader’s job changes first.

Move 5: Redesign the human system, then govern it separately.

Map the decision architecture explicitly: which decisions stay entirely human, which become human-AI collaborative, and which can be fully automated. What gets left to default will be decided by whoever deploys the next tool.

Then separate re-founding governance from the operating review entirely. Different metrics. Different time horizon. Protected investment that cannot be raided to hit a quarterly number.

Define the board’s re-founding indicators.

The risk that doesn’t make the board deck

The conversation in most boardrooms right now is about AI opportunity and AI risk. Often framed as technology questions with technology budget business cases.

But the real risk of this moment is not a failed implementation.

It is a successful one that optimizes you for a world that is being replaced underneath you. The boards and CEOs who will look back on this period as a defining win are not the ones who moved fastest on headcount. They are the ones who had the courage to ask the harder question: what are we actually re-founding this organization to become?

Most organizations I work with have an AI roadmap. Almost none have a re-founding plan.

What’s stopping yours?

Share

Subscribe now