Most of us were taught to treat disruption as an interruption to normal business. In 2026, the assumptions under your business can shift before the organisation has finished absorbing the last change, and that asks for a different kind of readiness from the people at the top.
Andy Demir | Qalbur Perspective | Leadership | 14 minute read
Executive Summary
The crisis-leadership model most boards inherited assumed that disruption interrupts normal business and then passes. That assumption is wearing thin. AI adoption, geopolitical fragmentation, labour scarcity, restructuring, and regulatory change are now converging, and in Deloitte's 2026 research, only 27% of respondents believed their organisations manage change effectively. For the people who run organisations, the task is shifting from preparing for the next shock to protecting the judgement, trust and management capacity the organisation will need for the one after it. My suggestion is to define readiness as the capacity to reconfigure repeatedly without exhausting what the next reconfiguration will require.
In This Article
- A Crisis Used to Have a Shape
- What the Older Crisis Literature Still Gets Right
- Europe Is Not Responding to the Same Problem in the Same Way
- Growth Can Be Disruptive Too
- France and Spain Are Treating Adoption as an Institutional Problem
- The Nordics Show What Happens After Adoption Stops Being Novel
- Central Europe Is Solving a Different Labour Equation
- The UK Shows What Happens When Change Outruns Absorption
- AI Is Making Organisational Slack Visible
- The Gulf Is Running a Different Experiment
- Trust Is Becoming More Expensive to Lose
- Three Capacities Now Matter at the Same Time
- What Senior Leaders Should Consider
- Questions for the Boardroom
- Readiness Is the Ability to Change Again
- References
A Crisis Used to Have a Shape
For a long time, crisis management carried an implicit sequence. Something abnormal happened: a major customer disappeared, a programme failed, financing tightened, a supply chain stopped or demand collapsed. The executive committee went into crisis mode, made the hard decisions, and expected to return eventually to something like normal. That mental model has become dangerous, because it teaches you to wait for the next clearly identifiable event. Today, several assumptions under your business can change before the organisation has absorbed the last one.
I am deliberately avoiding the phrase "permanent crisis". You cannot run an organisation indefinitely as an emergency room. Crisis language concentrates authority, tolerates exceptional workloads, shortens horizons, and, when made permanent, becomes an excuse for rushed decisions and weak design. The description I find most useful is "persistent discontinuity". A trade relationship becomes a security issue, a productivity tool becomes a workforce issue, a restructuring becomes a trust issue, and a skills shortage becomes a capacity constraint just as demand rises. ECB President Christine Lagarde made a similar point from the macroeconomic side at the Munich Security Conference in February 2026: the economic interdependence Europe long treated as a source of stability has become a source of vulnerability, both to global disruptions and to the deliberate weaponisation of dependencies.
Inside the organisation, the effect is cumulative, and most of it is invisible from the boardroom. Deloitte's 2026 Global Human Capital Trends survey found that only 27% of respondents believed their organisations managed change effectively, and only 8% felt their continuous learning needs were being met. A third of workers reported going through 15 major changes in a single year; 68% said their wellbeing had suffered and 60% that their workload had grown. Your company can survive every programme separately and still weaken as a system.
What the Older Crisis Literature Still Gets Right
The earlier version of this article leaned on an older book, and part of its argument survives. In August 2009, in the aftermath of the financial crisis, Stephen R. Covey, FranklinCovey chief executive Bob Whitman and Breck England, then a solutions architect at the firm, published Predictable Results in Unpredictable Times. FranklinCovey said the book rested on its study of more than 300,000 people in 17,000 work units across 1,100 organisations (the firm's own client research, not an independent study). The answer came as four principles: execute a small number of priorities with excellence, move with the speed of trust, achieve more of what customers actually value instead of cutting indiscriminately, and turn fear into engagement.
I would keep those principles. Focus, trust and the management of fear still decide performance. What I would leave behind is the assumption around them that unpredictable conditions are something an organisation passes through. The question for your board now is whether the organisation can change repeatedly without consuming the conditions that make further change possible.
Europe Is Not Responding to the Same Problem in the Same Way
There is no single European response to continuous change, because labour law, ownership, industrial mix and the supply of skills all shape what a leadership team can actually do. Germany shows the tension most clearly. At the end of September 2026, after its supervisory board had backed a plan to make the group markedly more profitable, Volkswagen gave notice on ten of its in-house collective agreements with effect from 31 December. Arne Meiswinkel, the board member for human resources at the Volkswagen brand, said the industry's economic environment had deteriorated sharply and the company wanted room to negotiate. IG Metall and the works council saw it differently: at the end of 2024 employees had accepted cuts worth around €1.5 billion a year in return for investment and job security until 2030, and they read the new demands as an attempt to reopen a bargain they had already paid for.
Both positions can be economically rational simultaneously, and that is where trust becomes an economic variable. Management may need more flexibility, and employees may reasonably believe that their earlier sacrifices bought a degree of security. If your board has asked a workforce for concessions in the last few years, the workforce is keeping its own ledger of that exchange, whether or not you are.
Germany also exposes a wider European trade-off. A European Commission working paper published in April 2026, using survey data from across the OECD, found a negative association between employment protection and the adoption of AI and other restructuring-intensive technologies, most pronounced among large incumbent firms. The opposite model has its own cost, because when restructuring becomes too easy an organisation can lose trust, experience and tacit knowledge faster than technology replaces them. So I would not set maximum flexibility as the goal. The goal is adaptability without institutional amnesia.
Growth Can Be Disruptive Too
The Netherlands shows a very different kind of discontinuity. ASML is dealing with strategic scarcity and extraordinary demand. In July 2026, the semiconductor-equipment maker reported second-quarter sales of €9.3 billion and net income of €2.9 billion, raised its full-year revenue outlook to €43–45 billion, and said it plans to add about 30% to its EUV and DUV immersion system capacity for 2027 as AI-related investment pushes its customers to expand faster.
Growth carries its own leadership risk, and boards usually see it later, because demand hides organisational weakness just as recession exposes it. Hiring accelerates, customers become more dependent, capacity expands, and investors raise their expectations inside one of the world's most geopolitically sensitive supply chains. The circumstances are the opposite of Volkswagen's, yet the question for the leadership team is the same: can you expand without letting today's scarcity lock in a cost structure, a supplier architecture, or a management system that will be hard to reverse when the cycle turns?
France and Spain Are Treating Adoption as an Institutional Problem
France has chosen state-supported diffusion over leaving AI adoption to individual firms. Its Osez l'IA plan, launched in July 2025 and reinforced in June 2026, aims, by 2030, for AI adoption by all large companies, 80% of SMEs and mid-sized firms, and half of microenterprises. In September 2026, the government reported that 615 AI ambassadors had made more than 35,000 companies aware of what AI could do for them. Awareness is the right word, since only around 70% of the plan's co-financed data and AI diagnostics were carried out in its first year.
The approach rightly treats diffusion as partly a capability problem, but its limits matter more to you. A government programme can provide access, training and incentives, but it cannot redesign work inside your company. It cannot decide which task an engineer should stop doing, or whether an employee gains more autonomy after automation or merely a higher target. Those decisions stay with the executive committee.
Spain has approached the problem from another angle. The IA360 plan that Prime Minister Pedro Sánchez presented on 21 September 2026 is a twelve-month roadmap that sets out technology and economic adoption alongside governance and what the government calls a social contract on AI among companies, workers, unions, parties, and public authorities. It is far too early to judge, but the sequencing is instructive for a board, because one of the recurring mistakes in corporate AI programmes is to separate deployment from the question of whether employees regard the resulting system as fair, useful and accountable.
The Nordics Show What Happens After Adoption Stops Being Novel
The Nordic countries move the question forward. OECD data show that in 2025 more than 35% of firms in Denmark, Finland and Sweden were using AI, against an OECD average of about 20%. Once adoption is normal, the hard questions become organisational. Which decisions stay with people? How do you measure performance when tools make some output much cheaper to produce? And how does a young engineer or analyst build expertise when the machine increasingly does the routine work through which expertise used to be acquired? Digital maturity leaves the management problems in place and changes which of them reach your desk first.
Central Europe Is Solving a Different Labour Equation
Poland and Czechia show why the same technology can lead to very different leadership decisions. The OECD's 2026 assessment of Poland describes record-low unemployment alongside the headwind of population ageing, and treats better skills in the working-age population as a condition for sustaining growth. Its most recent economic survey of Czechia reports an unemployment rate among the lowest in the OECD, persistent labour shortages, and an automotive sector closely tied to German demand, in which hybrid and electric cars accounted for more than a third of passenger car exports by 2023.
In that setting automation means something else. For a company facing persistent labour scarcity, AI may be less about removing people than about releasing scarce capacity; for a company facing weak demand, the identical technology looks mainly like a cost lever. Institutional environments shape the choices available, but the comparison is useful only when it brings you back to your own organisation and the consequences you will live with.
The UK Shows What Happens When Change Outruns Absorption
The United Kingdom adds a useful contrast, because companies there usually have more freedom to change structures and employment arrangements than in much of continental Europe. In Deloitte's 2026 UK findings, based on 345 UK responses, only 18% thought organisations managed change effectively, compared with 27% globally. Seventy-six per cent of UK workers said their wellbeing had suffered from too many changes at work, and 57% reported heavier workloads as a result of organisational change.
So be careful about equating decision speed with adaptability. A company can decide quickly and absorb badly. The next transformation starts while managers are still working through the last one. From the boardroom, the company looks agile; from the floors below, it looks exhausted.
AI Is Making Organisational Slack Visible
Most executive attention goes to what AI can save in hours, headcount and cost, and adoption is moving fast enough to make that urgent. The OECD reports that the share of firms using AI in member countries rose from around 7% in 2021 to 20% in 2025, and a 2026 study by Alexander Bick and colleagues found that 32% of workers in the large European economies it surveyed used AI in their jobs, against 43% in the United States. The productivity case usually stops one step too early, though. If a manager saves six hours a week, what happens to those hours? If the answer is automatically more output, you have captured an efficiency without necessarily making the organisation stronger. Give that manager five more direct reports and the coaching capacity disappears. When an experienced engineer finishes documentation faster, you can remove the apparent slack or reinvest some of it in mentoring, experimentation and problem-solving. That choice belongs to the executive committee and should be made deliberately, rather than by default, in next year's budget.
This is why I would distinguish spare capacity from waste. An unused hour is not always inefficiency, and an experienced employee whose utilisation falls below a benchmark is not necessarily surplus; some spare capacity is where organisations learn. The OECD's July 2026 paper Skills in the AI age supports this indirectly: it identifies skills shortages as a key constraint on adoption, and finds that managers, professionals and engineers are among the most exposed to AI yet less likely to be automated because their work relies on non-routine cognitive and social skills. The risk for your organisation goes beyond AI removing people. Executives may redesign the organisation around assumed productivity gains before anyone knows which tasks have really become cheaper, which capabilities have become more valuable, and where human judgement has become the new bottleneck.
The Gulf Is Running a Different Experiment
The Gulf's model of transformation is very different from Europe's. In April 2026, the UAE government announced that it would move at least half of federal government services and operations to agentic AI within two years. In May, the Cabinet approved the first package of services, a framework setting out each ministry's responsibilities, and a programme to train 80,000 federal employees across leadership, technical, specialist, general workforce and train-the-trainer categories. What interests me is the sequencing, because capability-building, governance, and organisational responsibilities are treated as part of the technology programme rather than being added afterwards.
None of this proves the UAE model is better. The speed of a decision proves nothing about implementation, and the programme is too new to judge. Moving so much government work to agentic systems also raises hard questions about accountability. The case is useful because it exposes the opposite trade-off to Europe's. Where institutional constraints are heavy, adaptation can become slow; where transformation is highly centralised, technology and structures may change faster than challenge, learning and the workforce can keep up. Neither condition guarantees better leadership.
Trust Is Becoming More Expensive to Lose
The older crisis literature was right to treat trust as operational, and in 2026 the argument has widened. The German edition of the 2026 Edelman Trust Barometer found that 74% of employees trusted their own employer, the only institution in Germany with a clear trust majority, while 81% of respondents were hesitant or unwilling to trust people whose values, information sources, approaches to problems or cultural backgrounds differed from their own. Edelman is a communications firm; this is survey research rather than an official statistic, and it describes Germany, not Europe. Even so, it raises a question you should take seriously: your company may be one of the few institutions where people still expect a coherent account of what is going on.
The duty that follows is narrower than it sounds: your people need to know what management knows, what it does not know, which evidence it is using, what has actually been decided and what remains uncertain. Repeated restructuring makes this harder, because every commitment made during one transformation becomes evidence employees use to judge the next. A broken commitment shows up later in slower execution and less discretionary effort, and your accounts will not isolate that cost.
Three Capacities Now Matter at the Same Time
It helps to separate three capabilities that are usually bundled under the term resilience. Operational resilience is the ability to keep functioning when something fails, whether a supplier, a system, a site or a market. Adaptive capacity is the ability to change when the assumptions behind the business change, reconfiguring roles, processes, skills and resources quickly enough to matter. Institutional coherence, the term I would suggest for the third, is the ability to change repeatedly without destroying trust, accountability, management capacity and organisational memory.
The first two are familiar to most boards, and the third is the one I would ask your board to examine more carefully. An organisation can be operationally resilient yet strategically rigid, or restructure quickly while losing the people who know why earlier decisions were made. Readiness in 2026 needs all three. That leads to the definition I would now use with more confidence than the one in the earlier version of this article: organisational readiness is the capacity to reconfigure repeatedly without exhausting the people, trust, judgement and institutional knowledge required for the next reconfiguration.
What Senior Leaders Should Consider
The first thing I would put in front of the executive committee is the total change load employees experience, rather than the number of transformation programmes. The CRM implementation, the AI pilot, the new reporting lines, the restructuring, and the cost programme may each fall under five different executive sponsors. For the person on the receiving end, they are one working environment, and nobody at the top is accountable for the sum. Someone should be, and the board should see that number next to the programme dashboards.
I would also treat management bandwidth as a capacity constraint, in the same way you treat plant capacity or working capital. Span of control tells you how many people report to a manager, but it says nothing about whether that manager has time left for coaching, judgement, customers, and the unexpected. When a cost programme removes a layer, ask what the remaining managers stopped doing to absorb the load; usually it is the work that builds the next generation of leaders.
Decide in advance, at executive committee level, what happens to AI productivity gains. Some should become lower-cost or higher-output, and some may be better reinvested in learning, customer contact, succession, and resilience, and finance should not automatically record that second category as a failure to capture savings. If the gains disappear into higher targets by default, you have still made the decision, only without discussing it.
Make human accountability explicit wherever AI supports decisions. Your people should know where machine assistance ends, who is authorised to rely on it and who owns the consequence when the answer is wrong. That is a governance question, so it belongs with the board.
Finally, name the parts of the organisation that should stay stable. Constant change is no virtue, and an organisation that keeps changing priorities because its leadership cannot decide is confused, however agile it looks. Selective adaptability is harder. It asks you to know what must change, what should remain stable, and which commitments deserve to survive even when the structure around them does not.
Questions for the Boardroom
- How many significant changes are our people absorbing at once, and which executive owns each programme?
- Which AI productivity assumptions have we already turned into higher targets or lower headcount before the gains were demonstrated?
- Where has our search for efficiency removed the management capacity or institutional knowledge we would need in a real shock?
- Which commitment from our last transformation are employees using to decide whether to believe this one?
- What part of the organisation should we deliberately keep stable over the next three years while the rest changes?
Readiness Is the Ability to Change Again
Under persistent discontinuity, the temptation is to reward speed above everything else. The organisation that restructures, automates or cuts first looks decisive, and sometimes it is. The harder test comes afterwards, when the assumptions shift again, and you find out whether the organisation still has enough trust, judgement, knowledge and management capacity to respond. That is why I would no longer define resilience as the ability to return to normal, since in many sectors there may be no stable normal to return to, nor would I define it as constant change, which can be another form of organisational failure.
The standard I would suggest instead is selective adaptability: knowing what must change and what must stay, and telling the difference before pressure makes the decision for you. That judgement cannot be delegated to a programme office; it rests with the board and the executive committee, and evidence that you are exercising it will not appear in the transformation dashboard. It will show up in whether your people still believe you the next time you ask them to change.
References
- ASML (2026). ASML reports €9.3 billion total net sales and €2.9 billion net income in Q2 2026. 15 July 2026.
- Banque des Territoires (2026). Le plan Osez l'IA veut passer de la sensibilisation à l'action. September 2026.
- Bick, A., Blandin, A., Deming, D., Fuchs-Schündeln, N. and Jessen, J. (2026). Mind the Gap: AI Adoption in Europe and the U.S. NBER Working Paper 34995.
- Covey, S. R., Whitman, B. and England, B. (2009). Predictable Results in Unpredictable Times. FranklinCovey Publishing. Research base as described in FranklinCovey's release of 7 August 2009.
- Deloitte (2026). Key UK findings from the 2026 Deloitte Human Capital Trends report, including global comparisons.
- Deloitte Romania (2026). Change fatigue: findings from the 2026 Global Human Capital Trends study.
- dpa (2026). Volkswagen kündigt fast alle Tarifverträge. 30 September 2026.
- Edelman (2026). Edelman Trust Barometer 2026: Deutschland.
- European Commission, Directorate-General for Research and Innovation (2026). Employment protection, adjustment costs, and technology adoption. Working document, 23 April 2026.
- Government of Spain, La Moncloa (2026). Pedro Sánchez announces the IA360 Plan. 21 September 2026.
- IG Metall (2026). Tarifrunde Volkswagen 2026. 29 September 2026, updated 30 September 2026.
- Lagarde, C. (2026). Preparing for geoeconomic fragmentation. Remarks at the Munich Security Conference, 14 February 2026.
- Ministère de l'Économie (2026). Un an du plan « Osez l'IA ». 3 September 2026.
- OECD (2025). OECD Economic Surveys: Czechia 2025.
- OECD (2026). AI use by individuals surges across the OECD as adoption by firms continues to expand. January 2026.
- OECD (2026). Foundations for Growth and Competitiveness 2026: Poland.
- OECD (2026). Skills in the AI age. OECD Artificial Intelligence Papers, 8 July 2026.
- UAE Cabinet, reported by Aletihad (2026). UAE Cabinet approves roles and responsibilities for Agentic AI project. 18 May 2026.
About the Author
Andy Demir is an international commercial and business leader with more than two decades of experience across complex B2B and industrial markets. Andy writes about executive leadership, strategy, organisational transformation, commercial performance, international business and the changing world of work, combining research with lessons drawn from operating across industries, markets and cultures.
