Many companies still talk about artificial intelligence as if it were simply the next generation of software—just another tool, another boost to productivity, another way to automate processes, cut costs, and speed up operations. This perspective is understandable, but it’s too narrow. AI does not merely change individual tasks. It transforms the economic logic that has organized companies for decades. This puts pressure on a level that has long been taken for granted: middle management.
For decades, middle management served as the backbone of large organizations. It linked strategy to operational execution, coordinated teams, consolidated reports, prepared decisions, and kept information flowing. Without them, many large companies would have been nearly impossible to manage. Their existence was no historical coincidence, but a response to a real problem: complexity. The larger companies became, the more difficult it was to gather, evaluate, and disseminate information and translate it into decisions. This required people. Many people. This need gave rise to hierarchies, reporting lines, steering committees, coordination roles, and management levels. This model made sense for a long time. It emerged in a world where information was scarce, slow, and difficult to evaluate. Those closer to relevant information had influence. Those who could structure it created value. Those who mediated between strategy and operational reality were indispensable. Middle management became the organization’s translation engine: distilling information upward, explaining it downward, and coordinating laterally.
But it is precisely this foundation that is currently shifting. Modern AI systems can analyze vast amounts of data in seconds, summarize project statuses, identify risks, consolidate communication flows, generate forecasts, identify resource requirements, and prepare decision options. Tasks that previously tied up significant management capacity are being automated, accelerated, or radically simplified. This changes more than just the efficiency of individual processes; it alters the economic necessity of many management functions.
AI reduces the cost of complexity
The public debate on AI often focuses on operational tasks, administrative work, or highly specialized expert roles. However, the greater structural pressure arises elsewhere: at the heart of the organization. This is where information is collected, sorted, evaluated, and passed on. Where meetings are prepared, coordination is organized, status reports are consolidated, and decisions are formulated. Where many roles are legitimized less by technical depth than by coordination, oversight, and mediation. It is precisely this work that relies heavily on knowledge processing. And it is precisely here that generative AI excels. It recognizes patterns, summarizes information, compares variants, identifies deviations, formulates options, and reveals connections. It does not automatically make the better decision. But it reduces the effort that was previously required to even reach a decision-making capacity. Therein lies the quiet devaluation of middle management. It is not that every manager is losing importance. But many traditional management tasks are losing their previous economic value. Those who primarily collect information, forward reports, moderate coordination loops, and organize operational oversight will find it harder in the future to explain why a separate management level is necessary for this.
The crucial question, therefore, is not: Can AI replace leadership? The better question is: Which aspects of leadership were never really leadership at all, but rather human infrastructure for slow information processing? Historically, complexity was expensive. The larger an organization became, the more coordination it required. The more markets, products, regions, and functions came together, the more management levels emerged. Each level had the task of reducing complexity, filtering information, and preparing decisions. This was necessary as long as information processing was slow and limited by human capacity.
AI is changing that equation. When information becomes available, analyzable, and prioritizable in real time, a significant portion of traditional management tasks becomes redundant. What used to require multiple meetings, Excel files, status updates, and management presentations can now be processed in significantly less time. Not perfectly and not without human review, but quickly enough to call old structures into question. This puts pressure on organizational models that have been considered professional for decades: long reporting chains, multi-level approval processes, complex matrix structures, extensive coordination routines, and meeting cultures that primarily keep the organization’s internal self-management running. Many companies have been optimized for controllability over the past decades. This made sense in more stable markets. Large organizations could afford to be slow as long as their market position was strong enough. Today, this slowness has become a risk. In an AI-driven economy, the demands for both decision quality and decision speed are rising simultaneously. Companies must recognize what is happening faster, assess what is relevant faster, and act faster. An organization that needs five levels, three committees, and eight coordination loops to do this does not lose because of a lack of technology. It loses because of its architecture.
What the numbers show
The figures show that this is not a marginal phenomenon. The McKinsey Global Institute estimates that Europe and the U.S. will each have to manage around twelve million career transitions by 2030. For Germany, McKinsey predicts up to three million career changes by 2030, which corresponds to about seven percent of total employment. Administrative office jobs are particularly affected; according to McKinsey, up to 54 percent of the expected job transitions in Germany could occur in this sector. At the same time, AI and other technologies could automate around 30 percent of the hours worked today by 2030, enabling a significant boost in productivity. [1]
These figures do not mean that millions of people will simply be replaced. That would be an oversimplification. The real shift runs deeper. AI is transforming work activities, task sets, and role dynamics. It makes certain work processes faster, cheaper, and more scalable. As a result, smaller teams can deliver results that previously required larger structures. A qualified employee with the right AI tools can prepare analyses, structure presentations, evaluate market information, summarize project statuses, and formulate decision options—tasks that used to keep several coordinators, analysts, or project managers busy.
In its index on occupational exposure to generative AI, published in 2025, the International Labour Organization also concludes that AI primarily transforms and complements tasks. Administrative and knowledge-based tasks are particularly affected. The ILO emphasizes that this often does not involve the complete replacement of entire professions, but rather the transformation of individual tasks within those professions. This is precisely what makes it so relevant for middle management: the role does not disappear overnight. Its core is gradually being eroded. [2] That is what makes this development so dangerous. Disruption rarely comes with a loud bang. It often begins quietly. First, a report is automated. Then an analysis. Then a summary. Then a planning draft. Then a status check. Then a basis for decision-making. Each individual change seems harmless. Taken together, however, the question shifts: Why is this role still needed in its current form?
Many executives underestimate this process because they initially view AI as a tool for support. It saves time, helps with writing, simplifies analysis, and speeds up routine tasks. That is true. But that is precisely where the ambivalence lies. A technology that makes people more productive also reveals which parts of their work previously consisted mainly of friction, repetition, and coordination. The better AI takes over these components, the more expectations shift toward the person behind the work. Support becomes a reassessment. Efficiency becomes a question of justification.
The Silent Crisis in Middle Management
This creates a massive tension for middle management. On the one hand, middle managers are expected to drive transformation, lead their teams, and integrate new technologies into the organization. On the other hand, they themselves are among those whose current roles are being threatened by these technologies. They are expected to facilitate a change that calls their own position into question. That is the psychological core of this crisis. The actual disruption is therefore not technical, but human. AI changes processes, but it also changes status, power, and self-perceptions. In many organizations, management has long been associated with an information advantage. Those who knew what was being decided at the top, who had access to relevant figures, and who could filter and distribute information held influence. As information becomes more widely available and easier to analyze, this advantage loses its significance. Power is shifting from control over information to decision-making authority. This is a fundamental change. Many management careers were built on the ability to reliably manage complexity. Those who coordinated, reported accurately, escalated risks, and mastered internal processes were considered good leaders. In a slower world, that made sense. In a faster world, it is no longer enough. Managing complexity is not the same as leading through complexity. One stabilizes the status quo. The other provides direction in the new.
Leadership is not going away. On the contrary: it is becoming more important. But it is changing. The leader of the future will not be judged by how many meetings they organize, how many reports they forward, or how well they master internal coordination processes. They will be judged by whether they create clarity, take responsibility, make decisions, develop people, and navigate uncertainty.
It needs to act less as a gatekeeper and more as a guide. Less of a controller, more of an enabler. Less of a conveyor of hierarchy, and more of an architect of effectiveness.
The World Economic Forum’s *Future of Jobs Report 2025* describes precisely this shift in the skills landscape. The report is based on assessments from more than 1,000 major employers worldwide, representing a combined workforce of over 14 million employees, and examines the expected changes in jobs and skills through 2030. Key future skills include analytical thinking, resilience, flexibility, leadership, social influence, and technological competence. This aligns with trends in management: it is not the role itself that remains stable, but only the aspect that generates real value in uncertain situations. [3]
What kind of leadership will create value in the future
This leads to an uncomfortable distinction. There are middle managers whose roles are being devalued by AI. And there are middle managers whose roles are becoming more important because of AI. Those whose work consists primarily of passing on information, monitoring, and administrative coordination are the ones being devalued. Those who turn information into direction, strategy into concrete action, uncertainty into decision-making ability, and change into learning ability will become more important. AI does not replace the ability to take responsibility. However, it is increasingly replacing the organizational overhead that was often confused with responsibility in the past. Companies must therefore take a very close look. The crucial question is not: How many leaders do we need? It is: What kind of leadership creates value in an AI-driven organization?
Anyone who asks this question honestly will realize that some layers have become too thick, some roles too vague, and some decision-making processes too slow. Many organizations did not build up their management structures because every level delivered strategic value. They built them up because complexity was virtually unmanageable without additional layers. If AI reduces this complexity, the old structure also loses some of its legitimacy. That does not mean that flat hierarchies are automatically better. That, too, would be too simplistic. Large organizations still need structure, accountability, and clear decision-making authority. But structure must no longer be an end in itself. Every level must be able to justify the value it creates—not historically, politically, or out of habit, but economically and operationally.
Does it speed up decision-making? Does it improve quality? Does it help people grow? Does it reduce risks? Does it foster customer engagement? Or does it mainly create more work for other internal roles?
At this point, AI becomes a mirror. It shows companies what can be automated. And it reveals how much work has accumulated within organizations because earlier systems were slow, fragmented, and cumbersome. Many meetings exist because information is inaccessible. Many reports exist because trust is lacking. Many coordination efforts exist because responsibilities are unclear. Many management roles exist because organizations could not manage their own complexity any other way. AI does not automatically solve these problems. But it makes them more visible.
Why AI alone isn’t enough
That is precisely why simply introducing AI as a tool is not enough. Many companies are currently making the same mistake: they are applying AI to existing structures and expecting transformation. But an old organization with new tools often remains nothing more than a faster version of the old organization.
If decision-making authority is unclear, even the best analysis is of little help. When leaders delegate responsibility upward, AI merely speeds up the preparation for the next round of coordination. If teams aren’t allowed to act, even the best basis for decision-making is of little use to them. Productivity doesn’t come from technology alone. It arises when technology meets the right structures.
The real challenge is organizational and cultural. Companies must clarify where decisions are made, which roles are truly needed, and how responsibility is distributed. They must distinguish between coordination that creates value and coordination that merely manages friction. They must evaluate leaders based on whether they generate impact, not activity. And they must be willing to rethink career paths. Because when traditional promotion logic relies primarily on span of control, titles, and hierarchical levels, it comes into conflict with a world where smaller, highly productive teams can achieve more than large, cumbersome structures. For many leaders, this is uncomfortable. In the future, it will no longer be enough to be a reliable part of the system. Those who want to remain relevant must demonstrate the difference they make—not through the number of meetings, the volume of reports, or the possession of information, but through judgment, clarity, courage, and the ability to empower others. This is more demanding than traditional coordination. But it is also more valuable.
At the same time, companies must not approach this development cynically. The devaluation of certain management functions is no justification for cold-hearted restructuring without accountability. On the contrary. Those who restructure organizations must bring people along with them, redefine roles, and take professional development seriously. Middle managers, in particular, often possess deep organizational knowledge, customer focus, an understanding of processes, and experience in dealing with people. This knowledge is valuable. The question is whether it remains trapped in old roles or is transformed into new forms of value creation.
Act now and win
Smart companies will therefore not simply cut middle management across the board. They will restructure it. They will ask which leaders can drive real transformation. Who empowers teams instead of controlling them. Who enables decisions instead of delaying them. Who integrates AI productively instead of treating it as a threat. Who delegates responsibility downward without losing direction. Who breaks down silos instead of managing them. These are the people we need. Perhaps more than ever.
The less astute companies will wait. They will initially treat AI as an efficiency initiative, roll out a few tools, organize training sessions, and hope that the rest will fall into place on its own. Then they will realize that while productivity may increase in isolated areas, the organization as a whole is not becoming more efficient. Because the real bottlenecks weren’t in the software, but in decision-making processes, power structures, and leadership paradigms. Only then will the economic pressure be great enough to truly address the underlying structure. For middle management, the situation is therefore serious, but not hopeless. The role isn’t necessarily going away. But it must redefine itself from the inside out. In the future, the value of a leader will lie less in passing on complexity and more in reducing it. Less in possessing information and more in creating meaning. It is not about continuing to exercise control, but about enabling responsibility.
Leadership thus becomes more entrepreneurial, clearer, and more effective. The key question for every middle manager is therefore: Does my value lie in providing genuine guidance, making decisions, and taking responsibility—or primarily in managing complexity?
This question is uncomfortable, but necessary. Because it is precisely this complexity that is currently being automated. For companies, the corresponding question is: Which parts of our organization are still designed for a world in which information moved slowly, coordination was expensive, and hierarchy was inevitable? And which of these structures are preventing us today from learning faster, making better decisions, and working more productively? AI forces organizations to take a more honest look at themselves. It reveals where leadership creates real value and where management merely manages the side effects of old complexity.
The plain truth is this: middle management isn’t simply undergoing a transformation. In many companies, it is currently losing its traditional economic foundation. But that is precisely where the opportunity lies. Those who redefine leadership can build organizations that are faster, clearer, and more human than the bureaucratic apparatuses of the past. Change isn’t just coming. It’s already here. The smarter companies are starting now to align their structures accordingly. The others will do so when the economic pressure is great enough.
written by Yasmin Floria Kreß – Business Partner of Frank Rechsteiner
Sources:
[1] McKinsey Global Institute: A New Future of Work: The Race to Deploy AI and Raise Skills in Europe and Beyond, 2024; and McKinsey Germany: AI Accelerates Disruption in the Labor Market: 3% Productivity Boost Possible, press release dated May 23, 2024 (https://www.mckinsey.com/mgi/our-research/a-new-future-of-work-the-race-to-deploy-ai-and-raise-skills-in-europe-and-beyond)
[2] International Labour Organization: Generative AI and Jobs: A Refined Global Index of Occupational Exposure, 2025. (https://www.ilo.org/publications/generative-ai-and-jobs-refined-global-index-occupational-exposure)
[3] World Economic Forum: The Future of Jobs Report 2025, 2025. (https://www.weforum.org/publications/the-future-of-jobs-report-2025)