Capability debt

Capability debt is what an organisation owes its own future when it automates the doing without redesigning the learning. Every time AI takes over a task that people used to perform, the work still ships, but the practice that built the underlying judgement quietly stops. Capability falls behind the tools, invisibly, because the outputs still look fine. Nobody sees a problem on any dashboard. The bill only arrives later, and all at once, when a decision turns up that the AI cannot make and no human in the room has been trained to make either. Like financial debt, it is borrowing against the future. Unlike financial debt, most organisations do not know they are taking it on.

Where the name comes from

The term borrows deliberately from technical debt, which Ward Cunningham coined in 1992 to explain why hastily shipped software would need rewriting: it is fine to borrow against the future, he argued, as long as you pay it back, and if you do not, you pay interest in the form of everything the shortcut makes harder later. Capability debt is the same idea moved from the code to the people. But the move changes everything about how the debt behaves.

Technical debt is visible to the engineers who carry it and repayable by refactoring. Capability debt is neither. It hides inside capable-looking output, so the people accruing it often cannot see it, and it cannot be cleared in a sprint, because rebuilding human judgement takes years of the very practice that was automated away. It is a slower, quieter and more expensive debt than the one it is named after, and it does not appear on a balance sheet.

What causes it

One thing causes capability debt: automating work faster than you redesign how people learn. Underneath that are three mechanisms, each with its own page in this research. The missing rungs are the junior tasks that used to carry people up to senior judgement, removed by automation before anyone noticed they were load-bearing. The missed reps are the repetitions handed to the machine, so the work is done but the practice never happens. And synthetic seniority is the surface effect: output that looks like the product of judgement the person has not actually built.

This is a drift problem before it is anything else. No leadership team decides to hollow out its own capability. It accumulates by default, one automated task at a time, because the tool arrives faster than the redesign of how people develop. Capability debt is what drift costs, measured in people.

The evidence for the underlying phenomenon

Capability debt is a synthesis rather than a single finding, but the components are well evidenced. EY's 2025 Work Reimagined survey, covering 15,000 employees and 1,500 employers across 29 countries, found that 88 percent of employees now use AI at work but only 5 percent use it in genuinely transformative ways, while 37 percent worry that overreliance on AI could erode their skills and expertise, rising to 43 percent in the UK. Crucially, EY found that organisations trying to capture AI gains on fragile talent foundations forfeited up to 40 percent of the productivity available to them. The tool without the human foundation does not deliver the gains; it delivers the appearance of gains while the foundation weakens.

The distributional evidence is just as telling. Brynjolfsson, Li and Raymond, studying 5,179 support agents, found AI raised the productivity of the newest and least experienced staff by 34 percent, while barely moving the most skilled, because the tool transfers expert patterns to novices. That is a real gain, and also the exact moment capability debt is created: the novice ships expert-looking work without the experience that expert-looking work used to require. The BCG and Harvard jagged-frontier experiment showed where the bill lands. On tasks just beyond the model's competence, consultants using AI performed worse than those with none, because they leaned on capability that neither they nor the tool actually had. And Gerlich's 2025 study connects the mechanism to the individual: frequent AI use correlated with weaker critical-thinking scores, mediated by cognitive offloading. Across all of it, the World Economic Forum's 2025 Future of Jobs report names skills gaps as the single biggest barrier to business transformation over the next five years. The debt is already being felt; it simply has not been named or measured.

How to recognise it

Capability debt rarely announces itself, so leaders have to look for its symptoms rather than wait for a number. The output is consistently good, but fewer people can explain how it was produced or defend it under questioning. Juniors cannot do unaided the thing the AI now does for them, and are not expected to try. Important decisions have no clearly accountable human owner, because the recommendation came from a system. Verification is treated as a rubber stamp rather than skilled work. And the organisation has quietly lost the ability to say which of its capabilities live in its people and which now live only in its tools. Any one of these is a sign that the debt is accruing.

Where the evidence remains uncertain

There is no settled way to measure capability debt directly yet, and honesty requires saying so. The evidence above establishes the components, deskilling risk, offloading, the novice-expert transfer, the productivity forfeited on weak foundations, but not a single validated index of the debt itself. It operates on a multi-year horizon that current workplace studies have not had time to track, and it is genuinely hard to separate from the real, immediate gains AI delivers. Building a proper measure is exactly why the SuperSkills work is developing a Capability Debt Index rather than asserting a figure. The prudent reading is that the phenomenon is real and under-measured, which is an argument for designing against it now, not for waiting.

How to reduce it

You reduce capability debt the way you avoid any debt: stop taking it on without deciding to, and pay down what you have. In practice that means four things. Decide in advance where human judgement must remain, rather than letting automation settle it case by case. Protect the practice that builds judgement: keep some work unaided, and build new development rungs on purpose to replace the ones automation removed. Treat verification as real, skilled work and staff it accordingly, because in an AI-assisted organisation the checking is the judgement. And measure capability directly, through live decisions, simulation and the ability to explain and defend work, rather than trusting that good output implies a capable person. This is the practical content of drift versus design, applied to the one asset that does not show up on the balance sheet.

Key research and primary sources

• EY (2025). Work Reimagined Survey 2025. 15,000 employees and 1,500 employers, 29 countries.
• Brynjolfsson, E., Li, D. and Raymond, L. (2023). Generative AI at Work. NBER Working Paper 31161.
• Dell'Acqua, F. et al. (2023). Navigating the Jagged Technological Frontier. Harvard Business School and BCG.
• Gerlich, M. (2025). AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking. Societies, 15(1), 6.
• World Economic Forum (2025). The Future of Jobs Report 2025.
• Cunningham, W. (1992), on the origin of the technical-debt metaphor: Introduction to the Technical Debt Concept, Agile Alliance.

Related SuperSkills research

Capability debt is the organisational accumulation of effects developed elsewhere: AI and human judgement, AI and critical thinking, the missing rungs, the missed reps, synthetic seniority and drift versus design.

About this research

Rahim Hirji is the author of SuperSkills: The Seven Human Skills for the Age of AI (Kogan Page, 2026) and the founder of The SuperSkills Intelligence Company. This work draws on research across more than 200 organisations in 30 countries over seven years. Findings are attributed to the studies that produced them and kept separate from the interpretation and the named concepts, which are the author's. Capability debt is part of the SuperSkills lexicon. This is a living reference, reviewed and updated as significant new evidence appears.