When Disruption Becomes the Operating Environment

byAndy Demir -

The traditional model of resilience assumed that disruption interrupted normal business. In 2026, organisations increasingly have to adapt as several assumptions change at once. That requires a different kind of readiness.

Andy Demir | Leadership Under Pressure | Leadership | 16 minute read

Minimalist Qalbur editorial cartoon showing an executive beside an organisation vehicle whose spare tyre has been removed as an “Efficiency Saving”, while a road sign points towards the “Next Transformation” and a pothole ahead.

Executive Summary

The old crisis-leadership model assumed that disruption interrupted normal business, and that assumption is becoming less useful. In 2026, organisations are dealing simultaneously with AI adoption, geopolitical fragmentation, labour scarcity, restructuring, regulatory change, and pressure to improve productivity. The leadership challenge is therefore shifting from preparing for the next shock to preserving judgement, trust, and organisational capacity while the organisation itself continues to change. I would define readiness today as the ability to reconfigure repeatedly without exhausting the people, management bandwidth and institutional knowledge required for the next reconfiguration.

In This Article

A Crisis Used to Have a Shape

The crisis management most executives learned carried an implicit sequence. Something abnormal happened: a customer disappeared, a programme failed, financing tightened, a supply chain stopped functioning or demand collapsed. Leadership moved into crisis mode, priorities narrowed, difficult decisions followed, and eventually the organisation expected to return to something resembling normality. That mental model is now becoming dangerous, because it encourages leaders to wait for the next clearly identifiable event. The harder reality is that several assumptions underneath the business can change before the organisation has finished adapting to the previous change.

I am deliberately avoiding the phrase "permanent crisis". Organisations cannot be run indefinitely as emergency rooms. Crisis language concentrates authority, tolerates exceptional workloads, shortens decision horizons, and, if used permanently, can become an excuse for rushed decisions and weak organisational design. The description I find more useful is persistent discontinuity. A trade relationship can turn into a security issue, a productivity technology into a workforce issue, and a skills shortage into a capacity constraint just as demand rises. The European Central Bank made a similar point from the macroeconomic side in February 2026, when Christine Lagarde told the Munich Security Conference that economic interdependence, long seen as a source of stability, had become a source of vulnerability, both to global disruptions and to the deliberate weaponisation of dependencies.

The organisational consequence is cumulative: a company can survive each programme separately and still weaken as a system. Deloitte's 2026 Global Human Capital Trends research, based on responses from more than 9,000 business and HR leaders across 93 countries, found that only 27% of respondents believed their organisations managed change effectively, and only 8% believed their employers met their continuous learning needs. A third of workers reported 15 or more major changes in the previous year, and 68% of respondents reported reduced wellbeing and 60% a heavier workload as a result of change.

What the Older Crisis Literature Still Gets Right

One older management source underpins some of my original thinking on this subject, and it is worth retaining only because part of its argument still holds. In 2009, in the aftermath of the global financial crisis, Stephen R. Covey, Bob Whitman and Breck England published Predictable Results in Unpredictable Times. Covey was already widely known for his work on personal and organisational effectiveness, Whitman was FranklinCovey's chairman and chief executive, and England designed the firm's effectiveness programmes for its clients. Their question was practical: why do some organisations continue to perform when economic conditions become highly uncertain?

The book drew on FranklinCovey's analysis of some 17,000 teams comprising more than 300,000 people, and its argument came down to four disciplines: execute a small number of priorities with excellence, keep trust high with every stakeholder, create more value instead of cutting indiscriminately, and stop fear from destroying engagement. I would keep those principles. Focus still matters, low trust still slows decisions, and fear still changes how people behave. Cutting the capabilities customers actually value can still improve a quarter while damaging the next three years.

What I would not carry forward is the surrounding assumption that unpredictable conditions are something an organisation passes through. That was understandable in 2009. It is much less useful in 2026, when the harder question is whether an organisation can change repeatedly without gradually consuming the conditions that make further change possible.

Europe Is Not Responding to the Same Problem in the Same Way

It is tempting to talk about a European response to continuous change, but there is no single one. Labour law, ownership structures, industrial mix and the supply of skills all differ from one country to the next, and so does the degree to which authority is concentrated or negotiated. Those differences shape what leaders can actually do.

Germany shows the tension clearly. In September 2026, IG Metall invoked a clause in its December 2024 agreement with Volkswagen to press management to recommit to the agreement, amid concerns about a new wave of planned cuts. At the start of October, the company gave formal notice to terminate several collective agreements at the end of 2026, citing cost pressures and changing economic conditions. The 2024 agreement had itself been a bargain: employees accepted no direct pay rises in exchange for protection from compulsory redundancies until 2030. Management may well need more room to act, and employees may well believe that their earlier sacrifices bought them a degree of security. Both positions can be economically rational at the same time.

That is why trust becomes an economic variable. When one side treats the issue as an economic matter and the other as a broken commitment, the next transformation becomes harder before it has even begun. Germany also exposes a wider European trade-off. A European Commission working paper published in April 2026 found, across OECD countries, a negative association between employment protection and the adoption of AI and other restructuring-intensive technologies, most pronounced among large incumbent firms. When workforce adjustment is expensive, technology adoption slows. The opposite model carries its own cost, because if restructuring becomes too easy, organisations can destroy trust, experience and tacit knowledge faster than technology replaces them. The objective I would set is adaptability without institutional amnesia, which is a harder target than maximum flexibility.

Growth Can Be Disruptive Too

The Netherlands shows a very different form of discontinuity. ASML is dealing with strategic scarcity and extraordinary demand. In July 2026 the Dutch semiconductor-equipment company reported €9.3 billion in second-quarter sales and €2.9 billion in net income, raised its full-year revenue outlook to €43–45 billion and said its order intake had remained very strong, with investment in AI-related chip capacity driving demand.

Growth of that kind creates its own leadership risk, because demand can hide organisational weakness just as recession can expose it. Hiring accelerates, customers become more dependent, capacity expands and investors adjust their expectations upward, all while ASML operates inside one of the world's most geopolitically sensitive technology supply chains. The management challenge differs from Volkswagen's, but the underlying question is related: can the organisation expand without letting today's scarcity lock in a cost structure, supplier architecture or management system that becomes difficult to reverse later? Resilience is tested during booms as well as downturns.

France and Spain Are Treating Adoption as an Institutional Problem

France offers another response: state-supported diffusion instead of leaving AI adoption entirely to individual firms. Its Osez l'IA plan, launched in July 2025 and expanded in June 2026, is designed to spread the use of AI across the productive economy. In September 2026, the French government reported that more than 35,000 companies had been made aware of AI through a network of 615 AI ambassadors in 20 regions and 13 sectors, and its 2030 targets are adoption by every large company, 80% of SMEs and mid-sized firms and half of microenterprises. The strength of the approach is that it treats technology diffusion partly as a capability problem, particularly outside large corporations with their own AI teams and investment budgets.

Its limitation is equally important. Government programmes can provide access, training, infrastructure and incentives, but they cannot redesign work inside the company. They cannot decide which task an engineer should stop doing, which management layer should change, or whether an employee gains more autonomy after automation or merely a higher target. That remains leadership work.

Spain has approached the problem from another direction. Its IA360 plan, presented by Prime Minister Pedro Sánchez on 21 September 2026 as a twelve-month roadmap, combines technology adoption and industrial ambition with explicit attention to governance, trust, and a proposed social contract on AI among companies, workers, unions, and public authorities. I find that sequencing instructive. One of the recurring mistakes in corporate AI programmes is to separate technology deployment from the question of whether employees regard the resulting system as fair, useful, and accountable, which is how a technically successful system ends up with failed adoption.

The Nordics Show What Happens After Adoption Stops Being Novel

The Nordic countries provide another point of comparison. OECD data show that more than 35% of firms in Denmark, Finland and Sweden were using AI in 2025, against an OECD average of about 20%. In the ICT sector, the OECD-wide figure was 57%, and in Sweden and Finland it was close to saturation at roughly 88% and 80%. This moves the leadership question forward. When adoption itself is no longer unusual, the difficult questions become organisational: which decisions remain human, how performance should be measured when tools make some output much easier to produce, and how somebody develops expertise when the machine increasingly performs the routine work through which expertise used to be acquired.

Labour Scarcity in Poland and Czechia

Poland and Czechia illustrate how the same technology can lead to very different leadership decisions across Europe. Both economies are deeply integrated into European industrial supply chains and are operating in tight labour markets. On Eurostat's measure, unemployment in May 2026 was 3.1% in Poland and 2.9% in Czechia, among the lowest rates in the EU. The OECD's 2026 assessment of Poland names population ageing as the main headwind and better skills in the working-age population as part of the answer if productivity convergence is to continue. Its 2025 survey of Czechia describes persistent labour shortages and buoyant wage growth in an economy whose car industry depends heavily on German demand, which accounts for about a fifth of the sector's value added, and is now undergoing the shift to electric vehicles.

This changes the meaning of automation. In an organisation facing persistent labour scarcity, AI can free up scarce capacity; in a company facing weak demand, the same technology will mostly serve as a cost-reduction mechanism. Geographical comparison is useful only when it leads back to the organisation. Countries do not manage companies: institutional environments shape the available choices, organisations respond to those choices differently, and leaders then have to live with the human and commercial consequences.

The UK Shows What Happens When Change Outruns Absorption

The United Kingdom adds another contrast, because organisations there often have more freedom to alter structures and employment arrangements than in parts of continental Europe. That flexibility can support experimentation, and it can also produce churn. Deloitte's UK findings for 2026 are striking. Only 18% of UK respondents believed organisations managed change effectively, against 27% globally; 76% of UK workers said their wellbeing had suffered from being tired of too many changes at work, and 57% reported heavier workloads as a result of organisational change.

Executives should be careful about equating decision speed with adaptability. A company can make decisions quickly and absorb them badly. A new organisational chart can be implemented while managers and employees are still occupied with the previous structure, and another transformation begins before the first has become routine. The company then looks agile from the boardroom and exhausted from inside.

AI Is Making Organisational Slack Visible

A great deal of executive attention is concentrated on what AI can save: hours, headcount, transaction costs, analysis time and service effort. Those are legitimate questions, and the scale of use makes them urgent. A Joint Research Centre survey of more than 70,000 EU workers, published by the European Commission in October 2025, found that 30% were already using AI tools at work, and OECD data show the share of firms using AI rising from 8.7% in 2023 to 20.2% in 2025. The productivity discussion, though, often stops one step too early.

If a manager saves six hours a week through automation, what happens to those six hours? If the answer is automatically more output, the organisation has captured an efficiency without necessarily strengthening itself. If that manager receives five additional direct reports, coaching capacity may disappear. If an analyst produces twice as much material, somebody else has to review it. And if an experienced engineer finishes documentation faster, the organisation can either remove the apparent slack or reinvest some of it in mentoring, experimentation and problem-solving.

This is why I would distinguish spare capacity from waste. An unused hour is not automatically inefficiency, and an experienced employee whose utilisation falls below a benchmark does not automatically represent excess capacity; some of that capacity is where organisations learn. OECD research supports the argument indirectly. Its July 2026 paper, Skills in the AI Age, identifies skills shortages as a key constraint on adoption and finds that high-skill occupations such as 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 leadership risk is therefore not only that AI removes people. It is that executives redesign organisations around assumed productivity gains before they understand which tasks have actually become cheaper, which capabilities have become more valuable and where human judgement has become the new bottleneck.

The Gulf Offers a Different Experiment

The Gulf provides a valuable point of comparison because its transformation model can be very different from Europe's. In April 2026 the UAE government announced a plan to move 50% of federal government services and operations to agentic AI models within two years. In May the Cabinet approved the first phase of services to be transitioned, a framework assigning responsibilities to ministries and federal entities, and a programme to train 80,000 federal employees in agentic AI, from ministers and senior executives to new joiners. What interests me is the sequencing: capability-building, governance and organisational responsibilities are being treated as part of the technology programme, where many organisations would add them afterwards.

That does not prove the UAE model is better. Speed of decision is not evidence of quality of implementation, and the programme is too new to judge by outcomes. Moving such a large share of government activity towards agentic systems may itself create difficult questions about accountability, employees' dependence on automated systems and the quality of decisions that become partially autonomous. The case is useful because it exposes a different trade-off. European organisations can face so many institutional constraints that adaptation becomes slow, while highly centralised transformation carries the opposite danger: technology and organisational structures may change faster than challenge, learning and workforce adaptation can keep pace.

Trust Is Becoming More Expensive to Lose

The older crisis literature was right to treat trust as operational rather than decorative, and in 2026 the argument has become broader. The German edition of the Edelman Trust Barometer found that 74% of employees trusted their own employer, making employers 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. The source needs to be read for what it is: Edelman is a communications firm, and this is survey research, not an official statistical series. Even so, the pattern raises an organisational question, because the workplace may increasingly be one of the few institutions where people still expect a coherent operating reality.

Nobody needs executives to settle questions of politics or social truth. Employees do 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 credibility of the next. An organisation can survive a broken commitment financially and still pay for it later through slower execution, greater resistance and less discretionary effort, which is a real economic cost even when accounting systems cannot isolate it.

Three Capacities Now Matter at the Same Time

I find it useful to separate three capabilities that are often bundled under the term "resilience". Operational resilience is the ability to keep functioning when something fails: can the organisation continue to deliver when a supplier, system, facility, or market is disrupted? Adaptive capacity is the ability to change when the assumptions behind the business change, so that roles, processes, skills and resources can be reconfigured quickly enough. Institutional coherence is the ability to change repeatedly without destroying trust, accountability, management capacity and organisational memory. The first two are familiar; the third is the one I think executives now need to examine more carefully.

An organisation can be operationally resilient yet strategically rigid, or highly adaptive yet culturally destructive, and it can restructure quickly while losing the people who know why previous decisions were made. Readiness in 2026 requires all three, which leads me to a definition I would use more confidently today than the one in my earlier article: organisational readiness is the capacity to reconfigure repeatedly without exhausting the people, trust, judgement and institutional knowledge required for the next reconfiguration. That is a more demanding standard than crisis preparedness, because it asks whether the organisation is preserving its ability to respond again.

What Senior Leaders Should Reconsider

The first thing I would measure is the total change load experienced by the employee, whatever the number of programmes on the executive dashboard. The CRM implementation, the AI pilot, the new reporting line, the restructuring, and the cost programme may fall under five different executive sponsors, but for the person receiving them, they constitute a single working environment.

Second, I would treat management bandwidth as a capacity constraint. Spans of control tell us how many people report to a manager; they do not tell us whether that manager has enough time left for coaching, judgement, customer discussion and unexpected problems. Third, I would decide in advance what happens to AI productivity gains. Some gains should translate into lower costs or greater output, and some may be better reinvested in learning, customer contact, innovation, succession, and resilience. Finance should not automatically classify the second category as a failure to capture savings.

Fourth, I would make human accountability explicit wherever AI supports decisions. Employees should know where machine assistance ends, who is authorised to rely on it and who owns the consequence when the answer is wrong. Finally, I would identify the parts of the organisation that should remain stable. Constant change is no virtue, and an organisation that keeps changing priorities because leadership cannot decide is confused, however adaptive it looks. Selective adaptability is harder: it requires leaders 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

  1. How many significant changes are our people absorbing at the same time, regardless of which executive programme owns them?
  2. Which AI productivity assumptions have already been converted into higher targets or lower headcount before the gains have been demonstrated?
  3. Where has our pursuit of efficiency removed the management capacity or institutional knowledge we would need in the event of a genuine shock?
  4. Which commitment from a previous transformation are employees using to decide whether they should believe the current one?
  5. What part of our organisation should remain deliberately stable over the next three years, even while the rest changes?

Readiness Is the Ability to Change Again

The executive temptation under persistent discontinuity is to reward speed above everything else. The organisation that restructures first, automates first or cuts first can appear decisive, and sometimes it is. But speed is only one part of adaptation. The harder test comes afterwards: does the organisation still possess enough trust, judgement, knowledge and management capacity to respond when the assumptions change again?

That is why I would not 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 use is selective adaptability: knowing what must change and what must remain stable, and being able to tell the difference before pressure makes the decision for it.

References and Sources

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.

About Andy →

Connect on LinkedIn →

Tags:

Related Reading

View all

Thanks for sharing your opinion!

Previous Post Next Post
Share to other apps
Copy Post link