In 2016, Wells Fargo was fined $185 million after it emerged that employees had opened more than 3.5 million fraudulent accounts without customer consent.1 The story made headlines as a scandal of corporate greed and individual misconduct, but beneath the headlines was a deeper truth.
Independent investigations revealed a workforce operating under relentless sales pressure, chronic fear of job loss, and an almost complete absence of psychological safety. People failed to uphold ethical behaviour not because they lacked values or capability, but because the conditions required to exercise good judgement had been systematically removed. The capacity for ethical, considered, high-quality work had been squeezed out of them long before any fraudulent account was opened.
Wells Fargo had invested heavily in targets, systems, and incentives. They had invested almost nothing in the human conditions those systems depended upon. When the gap between what it asked of its people and what it provided for its people became wide enough, the consequences were catastrophic.
In this series, we have argued that as AI drives down the cost of production, distinctly human capabilities become more valuable, not less: judgement, inquiry, attunement, presence, purpose, and accountability. We have also explored how the tools promising to free people up for that work may, in practice, be eroding the very capacity required to do it; increasing cognitive load, intensifying pace, and quietly wearing down the people organisations depend on most.
Which brings us to the harder question sitting beneath both of those arguments:
Can people actually make the leap?
The Wells Fargo story is not just a cautionary tale about corporate misconduct. It is a precise illustration of the gap this article is about: the difference between capability and capacity. Between what people are theoretically capable of, and what they can actually access in the conditions they are in.
The Strategy Built on Wishful Thinking
There is an assumption sitting at the centre of almost every AI workforce strategy right now, and it is rarely examined:
AI will handle the boring, repetitive work. Our people can now do more of the creative and strategic work. Let AI handle the process. People will handle the people.
The logic seems sound on the surface. Automate the routine, the procedural, the predictable, and humans are freed up for the creative, the relational, the complex. Organisations become more agile, more innovative, more human.
It is a comforting idea. It is also, in many cases, dangerously incomplete.
It skips over the real question: not whether people can do that work in theory, but whether they can do it in the conditions that actually exist.
For a significant proportion of the workforce, the answer is no. And these are not skills you can simply mandate through policy or check off in a half-yearly performance review. Judgement, inquiry, attunement, presence, purpose, and accountability are expressions of a person’s inner state. And inner states are shaped by conditions that most organisations have never considered within their remit.
Until we are honest about that, every AI workforce strategy built on the assumption of human readiness is a strategy built on sand.
The Gap Between Capability and Capacity
We are not facing a skills gap. We are facing a capacity crisis. They are not the same problem.
Capability is what people are able to do. Capacity is what they can actually access given the conditions they are in. AI is increasing the demand for the former while doing nothing for the latter. In fact, the evidence suggests it may be actively eroding it.
As we explored in a previous article in this series, research on cognitive load helps explain why. When high performers use AI tools on complex tasks, their mental effort does not decrease, it expands. The result is rising cognitive load, eroded clarity, and a constant tension between their own judgement and the machine’s. The prefrontal cortex, responsible for complex reasoning, ethical judgement, creative thinking, and long-term planning, is highly sensitive to sustained pressure. When the nervous system is under chronic load, activity there decreases. The brain shifts towards survival circuitry, towards the immediate and reactive.
Research from the University of California2 found that even moderate, sustained stress significantly impairs working memory, cognitive flexibility, and emotional regulation, precisely the capabilities organisations are now depending on humans to deliver. Critically, this degradation is largely invisible to the person experiencing it. People often believe they are reasoning clearly under pressure, but the evidence suggests otherwise.
Over time, heavy reliance on AI tools is associated with reduced independent critical thinking. The more people outsource or fragment their cognitive effort, the less practised they become at exercising it independently.
This is the double bind. AI is calling our distinctly human capabilities forward while simultaneously eroding the internal and environmental conditions required to deliver them. We may be building a workforce that is more dependent on human judgement and less able to access it under pressure.
Consider the variables that shape whether those conditions exist for a given person on a given day: their sleep, the state of their closest relationships, unresolved grief, chronic pain, financial stress, whether they feel safe with their manager, whether the culture rewards honesty or punishes it, whether they have had a single uninterrupted hour to think in the past week.
These are not peripheral factors. They determine whether capability on paper becomes capacity in practice, yet almost none of them appear in workforce strategy.
What Is Shaping the Capacity of Your People
The people in your organisation are not arriving as blank slates. They are arriving as full human beings, carrying lives that extend well beyond what most organisations consider their business. Mental health challenges affect, by conservative estimates, one in four people at any given time. Physical illness, exhaustion, grief, and relational strain are commonplace. The accumulated load of lives being lived under pressure does not begin nor end at the office door.
Moreover, organisational conditions often amplify these pressures. Gallup’s most recent State of the Global Workplace3 report found that 60% of employees are emotionally detached at work, with a further 19% describing themselves as actively miserable. A Deloitte Global Millennial and Gen Z Survey4 found that nearly half of younger workers have left a job due to burnout – not disengagement or for another opportunity, but due to exhaustion so complete that leaving felt like the only option. The World Health Organisation now classifies chronic stress and burnout as an occupational phenomenon5. This is not a marginal issue. A large share of the workforce is currently operating in a state fundamentally at odds with the higher-order work they’re increasingly being called on to do.
When we ask those same people to step into the most cognitively demanding, most nuanced, most deeply human work that exists – the work AI cannot touch – without addressing the conditions that determine their ability to do it consistently and sustainably, we are not being ambitious. We are being naive.
The Cost of Getting This Wrong
AI adoption without investment in human capacity does not elevate the workforce. It amplifies the fragility already within it. It accelerates output from people who are already stretched, increases the complexity of decisions being made by minds that are already overloaded, and raises the stakes of human judgement at precisely the moment the conditions for good judgement are most compromised.
Wells Fargo is the cautionary tale. AI at scale is the next chapter.
The organisations that will avoid repeating that pattern are not those with the most sophisticated AI strategy. They are those with the courage to ask a harder question than what skills do we need.
What do the human beings in our organisation actually need in order to think clearly, judge wisely, and connect genuinely over time?
And are we willing to provide it?
Human capacity does not emerge by default. It rests on conditions. On nervous systems that are not trapped in chronic threat. On teams where people feel safe enough to speak honestly and think clearly. On clarity, recovery, and structures that restore rather than continuously deplete. Not surface-level perks, but the conditions that make wise judgement, genuine connection, and sustainable performance possible. This means going beyond the wellness app or the early Friday finish, and building in protected, structured recovery that enables the brain and body to restore what demanding work depletes.
The real question is no longer whether your workforce can use AI. It is whether your organisation is creating the conditions in which human judgement, discernment, courage, and connection can still be accessed under pressure.
If the future you are designing depends on more distinctly human work, what are you doing to make that work truly possible?
And if you are not, what are you actually scaling?
