artificial intelligence (AI)

Design-oriented innovation research on AI applications

25.03.26

Around the world, many companies face the challenge of keeping pace with artificial intelligence (AI) and gaining a competitive edge through AI tools that leverage their specific expertise. This development is leading to a renaissance in design-oriented management and innovation research, which aims to bridge the gap between theory and practice. In the process, research is shifting from empirical work at universities to connective design in real-world laboratories of change.

In this blog post, I outline the development of the design-oriented research approach, highlight its advantages, and explain a general procedural concept.

A decisive year for AI adoption

2026 could be the year that determines for many companies whether they achieve a breakthrough in AI adoption or remain stuck with isolated pilot projects. According to an analysis by the market research firm Forrester, only 15 percent of decision-makers report that the use of AI has so far made a measurable contribution to their companies’ operating results. This could lead them to postpone about a quarter of their planned AI spending until 2027. However, the danger of being too hesitant in adopting AI is that it creates a gap that is difficult to close.¹

On the other hand, pioneering companies are succeeding in achieving significant competitive advantages with AI tailored to their specific situations. These pioneers practice design-oriented management and innovation research to lay the groundwork for success. One example is Siemens, which aims to develop the world’s largest and broadest range of industrial AI applications and double its enterprise value.²

The German Economic Research Institute (IW) forecasts that the potential for increasing gross value added through the use of AI in Germany by 2034 stands at 440 billion euros. Of this, 110 billion euros is attributable to potential innovations and 330 billion euros to increased productivity. For Germany as an industrial hub, the application of AI in particular presents an opportunity. Google, a subsidiary of the U.S. technology conglomerate Alphabet, has also recognized this. It plans to invest 5.5 billion euros in Germany and open a center for AI applications in Berlin, where its own researchers will also be based.³

This new situation has exciting implications for the design of AI ecosystems.

Designing AI ecosystems

The opportunity for established companies lies in knowledge-specific AI, where AI tools leverage and amplify the companies’ specialized expertise.⁴ This is most effectively achieved within AI ecosystems.

The term “AI ecosystem” refers to a dynamic network with strong connections between various actors who use AI technologies to create and disseminate innovations. AI ecosystems can be concentrated in a single region and have developed around a central hub, such as a university or a company. The world’s best-known and most influential AI ecosystem emerged in Silicon Valley in the San Francisco Bay Area. Stanford University near Palo Alto formed the core of Silicon Valley in the 1930s. In 1939, Bill Hewlett and David Packard founded the company HP there in a garage. Later, leading semiconductor companies – which gave the valley its name – and some of the major AI providers established themselves here.

Stanford University served as a model for Helmut Schöneberger, the head of the Munich-based startup incubator UnternehmerTUM. Crucial to its success were the collaboration with the Technical University of Munich and the support of BMW shareholder Susanne Klatten. In addition, other startup incubators have emerged in Munich, such as the Center for Digital Technology and Management (CDTM), which offers the transdisciplinary master’s program in Technology Management, works closely with industry partners, and has spawned over 250 startups.

By 2025, startups in Munich had received 3.3 billion euros in venture capital, thereby pushing Berlin into second place in this category, where startups raised 2.7 billion euros. However, AI ecosystems with successful startups have also emerged in other German regions.

Key players in AI ecosystems include:

Decision-makers in relevant policy areas and at various levels of state and regional government
Universities and schools
AI startups that originated as spin-offs or have established themselves in a region
Venture capitalists who use venture capital to finance the growth of AI startups
Major providers of AI hardware and software that shape financial markets with their financial clout and have a significant impact on the environment and society, as well as
established companies and their employees, who use AI, collaborate with startups, and invest in them.

Pharmaceutical research provides an interesting example of new opportunities for collaboration between established companies and startups.

In the pharmaceutical industry, tech-bio companies, such as the French startup Owkin, are accelerating early-stage drug discovery with the help of AI. A first wave of drugs developed in this way is currently in clinical trials. One possible form of collaboration is for AI biotechs to take on early-stage drug discovery and for established pharmaceutical companies to handle the later phases and commercialization. Another option is for pharmaceutical companies to acquire AI startups and build up their own relevant capabilities. In any case, this research-intensive industry is facing a fundamental change, in which the design of AI ecosystems is a key success factor.⁵ This example illustrates that as industries evolve, management research changes as well.

However, artificial intelligence is not only a driver of productivity and innovation but also poses a potential threat. In light of geopolitical shifts, it is crucial for Europe to regain its ability to shape the future. A trustworthy AI can be an important means to this end. The recent dispute between the U.S. government and the U.S. startup Anthropic illustrates just how political the design of AI ecosystems has become.⁶ This, too, is leading to significant changes in management research.

An important aspect that is often overlooked in AI ecosystems is the environmental and social impact of AI technologies. For instance, large language models require the work of crawlers that collect data and annotators who comment on, evaluate, and label texts and images. This work often takes place out of the public eye, frequently in countries of the “Global South.” Transparency in these supply chains is low.⁷

Management research in real-world laboratories of change

Empirical approaches, predominantly originating from universities, have long dominated business research. The authors publish the results of their work in academic journals, which practitioners rarely read. Critics of this retreat into an “academic ivory tower” argue that the practical relevance of research has diminished. This criticism is not new, but it has grown stronger.⁸

Universities defend their position by arguing that the practical relevance of empirical research is based on “field trips” in which the perspectives of relevant actors are analyzed. However, the goal of such work is generally not to design concrete objects – such as new AI-based business models –within their specific context.⁹

In this respect, business research differs from the design-oriented approach of engineering sciences, which takes place predominantly in laboratories and pilot plants.

Due to the growing importance of inter- or transdisciplinary research, the proportion of connective design is also increasing in management science. This practice- and design-oriented research is increasingly taking place in real-world laboratories of change.¹⁰ A key driver of this development is artificial intelligence, which acts as a catalyst for the founding of startups and the realignment of established companies. We believe that, following a phase of retreat into universities, management research should once again take place more strongly in practice and utilize design-oriented approaches.

This response to the questions of how and where research approaches should be conducted marks the beginning of a new chapter in the history of management research. A key feature of this reorientation toward connective design is research in which scientists, consultants, and practitioners bring their diverse perspectives and strengths together. The benefit lies in better solutions to complex problems based on new scientific insights.¹¹

The development of humanoid robots provides an example. In this rapidly growing market, companies have the best chances when they combine software and hardware expertise. Data from the real world is the biggest bottleneck here. The Metzingen-based manufacturer Neura Robotics has therefore decided to build training facilities for robots. One of the first of these “gyms” is being developed in collaboration with the Technical University of Munich at the Munich Institute of Robotics and Machine Intelligence (MIRMI) at Munich Airport. In international competition with companies from the U.S. and China, rapid scaling is crucial for this AI application.¹²

Since design-oriented management research is less widespread, I would like to briefly outline its history. Personally, over the past few decades, I have repeatedly observed how different the thought and language patterns of the relevant actors are in theory and practice.

Principles and pioneers of creative management research

Design-oriented management research is based on similar principles and has been shaped by several pioneers.¹³ These principles are:

Action research, pioneered by social psychologist Kurt Lewin. Lewin advocated for a connection between theory and practice to solve real-world problems.
General design theory, as defined by Nobel Prize-winning economist Herbert Simon.¹⁴ Simon defines this as the science of designing human-made artifacts and systems.
Action Science, conceived by Harvard professor Chris Argyris.¹⁵ Argyris’s primary goal is to make knowledge usable in order to improve actions within organizations.
The Design Science Research (DSR) approach developed by Alan Hevner at the University of South Florida.¹⁶ This concept, which is particularly widespread in business informatics, focuses on solving complex, real-world problems through innovative IT systems.
Action Design Research (ADR), described by various scholars, which combines Action Research and Design Science Research.

Pioneers of design-oriented management research include Joan Ernst van Aken and Georges Romme, who teach at Eindhoven University, as well as David Denyer from Cranfield University in the UK. Van Aken aims to bridge the gap between management theory and practice. The focus is on transdisciplinary research to solve practical management problems. Building on design theory, Romme tests iterative processes in organizational design.¹⁷ Denyer views management knowledge as a malleable resource for solving real-world problems. His solution-oriented approach seeks to ensure practical application through mechanisms of action and evidence-based verification using various information sources.

The advantages of a design-oriented approach are particularly evident in innovation and sustainability research.

Advantages in innovation and sustainability research

When applying AI, the advantages of a design-oriented approach to innovation and sustainability research unfold most notably in companies that practice this approach in self-similar Strategy 5.0 labs. The goal of such a real-world lab is to connect the various fields of action for AI applications.¹⁸ It is helpful if the countercurrent principle – combining a top-down AI strategy with a bottom-up, harmoniously diverse range of AI applications – is effectively implemented.

Based on our experience from a series of projects, the following benefits are of particular importance:

The removal of barriers to innovation
high-performing and trustworthy innovation ecosystems
motivated high-performance teams
connective design as a core competency
improved technology transfer
accelerated learning loops using agile methods, as well as
concrete results and measurable success for all stakeholders.

These advantages enable a realignment of innovation systems.

The shortcomings in Germany’s energy and mobility transition are an example of how important it is to remove barriers to innovation.¹⁹ The aim here is to translate new insights from innovation research into practical action.

The goal is to design high-performing and trustworthy innovation ecosystems.²⁰ A hallmark of these systems is improved collaboration among stakeholders from the political, scientific, economic, and social sectors.

Motivated high-performance teams play a key role in shaping innovation research.²¹ Leaders have the task of exemplifying and fostering an entrepreneurial mindset. This begins with education and continues throughout one’s professional career.

A core competency to be developed in this context is connective design.²² This requires a reorientation of teaching and research using AI as a tool. Such human-centric AI permeates all disciplines.

A positive side effect is improved technology transfer.²³ The goal here is to overcome the German paradox between strengths in basic research and weaknesses in commercialization.

One way to achieve this is through accelerated learning loops within organizations.²⁴ Agile methods are used for this purpose. It is important to adapt a general approach to specific types of problems and the situation at hand.

In doing so, all stakeholders should focus on concrete results and measurable successes.²⁵ In joint programs, this is achieved through transparent performance management.

In the following, I would like to explain a suitable approach.

Steps of a general procedural concept

An example of design-oriented innovation and sustainability research is the AI-based strategic and organizational realignment of a company.²⁶ For topics such as this, a general procedural concept consisting of the six steps shown in the figure has proven effective.

The first step is an analysis of the state of research, best-practice examples, and the specific initial situation. The task here is to combine an internal and an external perspective in an audit.

It is then crucial to develop a comprehensive understanding of the problem’s complexity, to identify its causes, and to coordinate efforts among the relevant stakeholders. Collaboration between various disciplines and practitioners is more successful than disciplinary research approaches.

Transdisciplinary teams also play a key role in the subsequent design and selection of creative solutions. What may be uncharted territory for a single organization can, when taken as a whole, contribute significantly to scientific progress.

A defining characteristic of design-oriented innovation research is the implementation of pilot projects for Minimum Viable Solutions (MVS). These pilots are tested in real-world laboratories of change. For established companies, working in learning loops often requires a shift in mindset. Human resources development can support this effort.

The fifth step is planning and executing the implementation. This is closely linked to financing the scaling process. German startups have long complained of disadvantages, for example, compared to the U.S. Therefore, an improvement in the political framework conditions should be sought, particularly in this step.

In parallel, transparent performance measurement takes place, for example, using the Objectives and Key Results (OKR) method. In this context, “transparent” means that performance management does not occur in sectoral silos. For a policy that sets goals but neglects to measure success, this involves a learning process aimed at joint system design by innovation managers.

Innovation managers design complex, evolutionary systems

As early as in our 2014 book publication “The Innovation Manager,” we concluded that a central task of innovation managers lies in the design of innovation systems and the connection of various fields of action.²⁷ Subsequently, we engaged intensively with the behavioral economic perspective of such connective design.²⁸

The scientific foundation for connective design is the theory of complex, evolutionary systems. The application of this approach to socio-technical systems has triggered a paradigm shift in strategic management.²⁹ U.S. digital companies have mastered this new management paradigm better than the European economy.³⁰

Another insight is that innovation research is a transdisciplinary design task.

Interdisciplinary or transdisciplinary?

We deliberately use the term transdisciplinary to make it clear that, unlike the term interdisciplinary, it is not merely about mediating between scientific disciplines, but also about involving non-scientific actors and establishing a connection between theory and practical design.³¹ The particular complexity of transdisciplinary innovation research stems from

the heterogeneity of the disciplines, which ranges from natural science and technical research through various policy fields to management science and organizational psychology
the diverse interests and thought patterns, e.g., of scientists and practitioners
the dynamics of development and the number of levels, from geopolitics down to the individual, as well as
the different roles of the actors, e.g., as neutral observers or personally affected individuals.

Unfortunately, transdisciplinary innovation research has so far lacked recognition in universities. The reasons for this are manifold. One important reason is likely that the traditional academic and publication system tends to reward disciplinary excellence. This presents an opportunity for applied research.

Transdisciplinary research on sustainability innovations

Our book, published in 1994, on the “ecological reorientation” of automotive companies emerged from consulting projects and accompanying research at the University of Stuttgart.³² This research was transdisciplinary, but had only a limited impact because German companies and policymakers at the time did not implement our recommendations – for example, regarding new propulsion systems.

Thirty years later, this industry – so vital to the German economy – is grappling with serious problems. VW, the world’s second-largest automaker by vehicle sales, is undergoing a process of strategic and organizational realignment.³³ At the same time, the geopolitical landscape is marked by extreme uncertainty.

This example illustrates that, when it comes to sustainability innovation, transdisciplinary and design-oriented research is of crucial importance for securing our country’s prosperity. ³⁴ In recent years, the field has evolved dynamically. The intersection of environmental technology and artificial intelligence has moved to the center of attention. New market leaders can emerge from collaborations between established companies and digital greentech startups if European policymakers succeed in improving the framework conditions.³⁵

One example is the German-Luxembourgish startup R3 Robotics, which has developed an AI-powered robotics platform for battery recycling. In this way, Europe can reduce its dependence on imports and increase the sustainability of batteries.³⁶

In my personal experience, students remain very interested in such topics because they recognize the resulting career opportunities. This provides important impetus for our expert network.

Combining management consulting and human resources development with design-oriented innovation and sustainability research

Our expert network Competivation has long combined management consulting and human resources development. The advantage for clients is a better value for money than with traditional consultants, as the focus is on the qualifications of the employees who are actively involved in the projects.

For several years now, we have been supplementing this service with design-oriented innovation and sustainability research. In doing so, we supervise the theses of dual-track students and external doctoral candidates who are working on relevant projects within the client organization. The advantage here is that the research is tailored to the specific situation of the company. The researchers and their companies benefit from the extensive experience of our experts.

With this approach, Competivation has created a unique international selling point geared toward the needs of the AI era.

Design-oriented research also has far-reaching implications for university teaching. We see ourselves as an innovative educational provider that teaches the ability to a of AI-supported connective design of solutions for complex management problems.³⁷

Conclusion

When applying AI, many companies face the challenge of defending their position and gaining competitive advantages through a specific approach
In this context, the design of AI ecosystems is crucial
Parallel to this development, research approaches and locations are shifting toward design-oriented innovation research in real-world laboratories of change
In this context, design-oriented management and innovation research is experiencing a renaissance stemming from a number of advantages
In this transdisciplinary research approach, a six-step procedure has proven effective, which is adapted to the respective problem type and situation.

Designing trustworthy high-performance systems

29.01.26

In recent years, the importance of connective strategic management has continued to grow. In light of the dynamic development of artificial intelligence (AI) and new geopolitical challenges, the design of trustworthy high-performance systems has become a focal point of interest. In this context, the term „high performance“ is being reinterpreted in business and politics. An important field of action here is design-oriented management research.

In our first blog post of 2026, I address the question of what important areas of action for Europe in terms of trustworthy high-performance systems are.

High performance in business and politics reinterpreted

For Jeanette zu Fürstenberg, who is responsible for Europe at the US fund General Catalyst, there is an opportunity for the old continent in connecting startups with the world of established industrial companies. Her successful investments include Mistral in France and the defense company Helsing in Germany. These companies focus on artificial intelligence (AI) that uses highly specialized application knowledge. Her publication „Wie gut wir sind, zeigt sich in Krisenzeiten“ (How good we are is revealed in times of crisis) was named Management Book of the Year in 2025. For her, the basis for a European high-performance system that can achieve reindustrialization is resilience, which enables recovery as quickly as possible after external shocks.¹ In 2025^, the number of startups founded in Germany reached a record high.

The topic of high-performance organizations is not new. High-performance organizations are characterized by high-performance teams. As early as the 1950s, the British Tavistock Institute developed an initial foundation with its socio-technical systems approach. I described the results of consulting projects on the characteristics of high-performance organizations in an article in Harvard Manager magazine in 1988. One important finding is that visionary leadership creates the framework for teams that work in a more self-organized manner.²

McKinsey consultants Jon Katzenbach and Douglas Smith examined the question of what characterizes high-performance teams.³ However, further developments have shown that, despite considerable efforts, empirical research is struggling with the design of high-performance organizations.⁴ AI is now giving performance management new impetus to improve the connection between strategy implementation and motivation.⁵

US tech companies with artificial intelligence (AI) now dominate the global economy. By the end of 2025, 61 of the world’s 100 most valuable companies will be from the US. The dominance of the US results from the unique strength of seven tech giants, which together have a market value of €18.3 trillion. An important driver of this development is the hype surrounding artificial intelligence (AI). Germany is represented in the top 100 ranking with the companies SAP (rank 40), Siemens (72), the European joint venture Airbus (91), and Allianz (100). In view of geopolitical changes, this concentration of power raises the question of how great the danger of dependence on the US is. ⁶ In the AI chip market, competitive pressure is increasing for market leader Nvidia.

AI chips are becoming increasingly powerful, but at the same time, AI increases the risk of disinformation. With a global market share of 85.2%, AI chip manufacturer Nvidia has a dominant position ahead of Broadcom (10.3%), Marvell (2.1%), and AMD (1.8%). Challengers AMD and Meta have announced a new AI system for data centers (Helios platform) that is expected to deliver a significant performance boost. Nvidia is countering with its new Rubin chip generation.⁷

However, there is a risk of a loss of trust in AI due to the risk of disinformation from fake accounts. AI bots falsify content, imitate people, and post automatically on social media. Such deepfakes can cause great economic damage and^, for example, ruin a brand’s reputation.⁸

Large language models and free AI tools often lead to a loss of quality and trust because they are not trained for high performance, but rather for the production of average knowledge. When AI users are under time pressure and there are no quality standards in place, „AI slop“ can result. Although this produces faster results, the quality declines. Possible consequences include a loss of reputation and trust. When using AI^, it is therefore important to supplement content with expert knowledge after quality control.⁹

AI and geopolitical challenges are reinterpreting the concept of high performance. Not all AI is trustworthy. We understand a trustworthy high-performance system to be a system (e.g., a company, a region, or a state) that performs very well compared to the competition and is trusted by the recipients of its services. In addition, these service recipients are willing and able to pay for the services. High-performance systems must therefore justify their higher prices (e.g., through „German quality,“ technical superiority, or a luxury brand).

Reinterpreting high performance means that high-performance systems are characterized by both success and trustworthy behavior. If neither of these is the case, we speak of system failure. Most socio-technical systems fall somewhere in between. Cases where only one of the two criteria is met are interesting. An existing pattern of success is at risk when a previously successful system, such as that of the AI champions, loses trust. This could result in a transitional phase with new opportunities, for example, if Europe, which has been less successful in digitalization to date, scores points with trust.

Peter Frankopan, a British professor of global history teaching at Oxford, sees the world in a transitional phase similar to that of the 1920s, when the old order was not yet dead and a new one had not yet been born.¹⁰

The question is therefore how Europe can seize its opportunities and become a designer of trustworthy high-performance systems.

Strategic realignment in a phase of transition

In the first nine months of 2025, DAX companies spent €6 billion on restructuring. The highest restructuring costs in 2025 were incurred by Mercedes (€1.4 billion), Volkswagen (€900 million), Siemens and Commerzbank (€500 million each). The automotive, mechanical engineering, and chemical industries are particularly affected. At the end of September 2025, 120,300 fewer people were employed in German industry than a year earlier. Many companies are offering generous severance packages. Often, one round of restructuring is followed by another without solving the underlying problems. This would require a strategic realignment after restructuring.¹¹

The term „strategic realignment“ describes an innovative approach to coordinating existing and new system elements (e.g., business model, strategy, technologies, customers, competencies, organization, culture, and environment). Realignments usually have a profound effect over a longer, undefined period of time in many parallel learning steps. Complex interactions play an important role in this process, resulting in specific patterns that are difficult to predict.

During a phase of transition, companies must manage complex realignment processes. In a successful, innovative company, important system elements are well coordinated. This alignment often takes place through fine-tuning, in which management continuously adapts the strategy to changes in the environment, for example. If this is not done, the company develops in the direction of misalignment. Management and supervisory boards often recognize this creeping decline too late. The result is an established company in a permanent crisis that requires restructuring.

The terms restructuring and transformation are now often used synonymously. Both terms describe a temporary, comprehensive change. Unfortunately, the inflationary use of the term transformation conveys the illusion that complex realignment processes are limited in time. The example of artificial intelligence clearly shows that such a static worldview is naive.

A longer-term goal of strategic realignments is the design of trustworthy high-performance systems.

Fields of action for trustworthy high-performance systems

In times of increasing polarization, high-performance systems are characterized by their ability to bring people together. History teaches us that the risk of polarization increases during periods of technological and political upheaval. This also applies to the changes brought about by artificial intelligence (AI). It is crucial that people see themselves as active participants in shaping change rather than passive objects of it. The complementarity of humans and AI is a malleable system. The performance of such a system depends on the ability to improve connections between the actors and the system elements. In our application-oriented research and teaching, we start with the thesis that the following fields of action, shown in the figure, are important in the design of trustworthy high-performance systems:

A connective strategic management for a triple realignment
high-performance teams with a growth mindset in a phase of transition
the connection of trustworthy partners from politics, business, science, and society, and
design-oriented management research in real-world laboratories of change.

Interdisciplinary university teaching faces the task of imparting the relevant skills for these fields, e.g., in the area of entrepreneurship for AI applications.

In the following, I will discuss these fields of action and skills in more detail.

Connective strategic management for a triple realignment

Since the 1960s, new challenges have led to various stages of development in strategic management.¹³ We distinguish between

a market- and finance-oriented stage (Strategy 1.0)
a technology- and innovation-oriented stage (Strategy 2.0)
a sustainability-oriented stage (Strategy 3.0) and
a resilience-oriented stage (Strategy 4.0).

In the current fifth stage of development (Strategy 5.0), the challenge lies in connecting the previous stages. Companies must become more resilient, more digital, and more sustainable at the same time.¹⁴ This requires the connective design of threefold strategic and organizational realignments. Such a triple realignment takes place in the context of serious changes in the political environment. The current situation is historically unprecedented. Therefore, the contextual intelligence of management plays an important role.¹⁵

German policymakers should create the framework for a cohesive strategic management approach through fundamental reforms. When the first signs of macroeconomic weakness appeared in 2018, they were harbingers of the most severe and longest industrial recession the Federal Republic has ever experienced. German industry has since lost much of its competitiveness. Experts are calling for new approaches to supply-oriented innovation policy and communication that conveys the need for a change of course. Politicians must implement the promised fundamental reforms. Such a new beginning can only succeed with solidarity instead of polarization.¹⁶

In this environment, resilience-oriented strategic management is becoming increasingly important.¹⁷ At this year’s World Economic Forum in Davos, the differing positions of the US president and European representatives clashed.¹⁸ Canadian Prime Minister Mark Carney suggests that in a world where major powers are becoming imperialists who blackmail other states, middle powers and smaller countries should form trustworthy partnerships.¹⁹

Such cooperation plays a decisive role not only at the geopolitical level, but also in high-performance teams.

High-performance teams with a growth mindset in a phase of transition

Black Forest Lab (BFL), currently Germany’s most valuable AI startup, is based in Freiburg, was founded in 2024, and develops AI models for image generation based on text. The founders are part of the core team behind the open-source AI model Stable Diffusion, the text-to-image model that generates digital images from text and, alongside ChatGPT, sparked the global AI hype in 2022. BFL’s Flux models are now one of Google’s biggest competitors. Important impetus for the work of the founding team came from Björn Ommer, a professor of computer science at LMU Munich. This example shows that high-performance teams can also emerge in Germany in the field of AI.²⁰

New ideas and the creation of something new often originate from people who find a state of flow motivating. The term flow (in the sense of „being in the flow“) was coined by psychology professor Csikszentmihalyi back in 1975. It refers to being completely absorbed in an activity, which usually involves a high level of intrinsic motivation and a change in the perception of time. Interviews in which outstanding creative personalities from various fields look back on their working lives show that their motivation stems primarily from the creative process. For many people, the foundations for possible flow states are often laid in their youth, based on their growth mindset.²¹

In her book Growth Mindset, Stanford professor Carol Dweck distinguishes between a static and a growth mindset.²² The following figure compares these two mindsets. High-performance systems often have leaders with a growth mindset. An important characteristic is that these people are aware of their talents, but place greater emphasis on their further development and learning processes. In contrast, people with a static worldview place greater hope in the effect of their innate talents.

Microsoft CEO Satya Nadella writes that the book had a strong influence on his personal development.²³

A person’s self-image and their environment are closely linked. High performance therefore arises from an interplay between the two. Social psychologist Mary Murphy has extended the growth mindset concept to organizations, their culture, and the environment surrounding them:²⁴

According to this, a growth culture promotes the potential of all employees. This culture emphasizes collaboration, continuous learning, and the development of skills.
A genius culture, on the other hand, believes in innate talent. This leads to internal competition, risk aversion, and a reluctance to admit mistakes.

Recommendations for action for managers are

create psychological safety and
giving constructive feedback.

However, simplistic application of this approach in practice underestimates the complexity of implementation. This can lead to demotivation among exceptional talents.

This raises the question of whether there are any current examples of a growth culture in Germany. A new bridge in South Westphalia has become a symbol of connective design. The Sauerland motorway is the most important transport link between the Ruhr area and Frankfurt. Due to the risk of collapse, the Rahmedetal bridge near Lüdenscheid, where I grew up, had to be suddenly closed in December 2021 and later blown up. This was a disaster for the economy with its many hidden champions and for the people in the region. Every day, 20,000 vehicles had to be diverted via bypasses and through residential areas. The German Economic Institute estimates the damage to businesses at around 1.5 billion euros. In Germany, new construction normally takes around eight to ten years. However, traffic is already rolling across one side of the A45 bridge via the after a record-breaking four years. This was made possible by smooth cooperation between the parties involved, a new planning procedure, and innovative construction methods. The German Chancellor sees this as a model for other renovations, and for the Minister President of North Rhine-Westphalia, the new benchmark for implementation speed in Germany is called „Rahmede“.²⁵

We can therefore summarize that the culture of socio-technical systems is strongly influenced by the mindset-image of important stakeholders and prevailing design patterns.

High-performance cultures are characterized by a growth mindset and a connective pattern. The opposite is a silo culture or, in extreme cases, a culture of self-satisfaction. Here, a static mindset and a divisive design dominate. Descriptions of outstanding leaders often heroize a lone wolf culture. These individuals are attributed with a growth mindset. At the same time, however, the impression is created that their successes were achieved single-handedly and in isolation from others, which is usually not the case. Until a few years ago, a culture of complacency was widespread in Germany. People rested on the successes of the past, but the mindset in politics and business was rather static and not very future-oriented.

Managers serve as role models in this regard. Their growth mindset is transferred to their employees. Conversely, managers with a static mindset and isolating behavior are responsible for the emergence of toxic cultures. Their position of power enables them to oust internal competitors and employees with a growth mindset that they perceive as a threat. Unfortunately, the role of consultants is often to secure and expand the position of power of the „static“ individuals. Attempts by external parties to change silo cultures are usually met with rejection and fail. It is therefore the task of supervisory bodies to review the dysfunctional mindsets of managers and take timely action. If this does not happen, there is a risk of system failure.

On its way to becoming a high-performance system, Europe currently finds itself in a difficult situation.

Connection of trustworthy partners from politics, business, science, and society

Europe first needs a resilience program against its enemies from outside and within. US political scientist Francis Fukuyama believes that Trumpism will continue even without Trump. For open democratic societies, this is an extremely dangerous development. He fears a relapse into the world order of the 19th century. It is therefore important that Western societies develop sufficient resilience. It should also be taken into account that tech billionaires primarily act in their own economic interests. The greatest danger for Europe is resignation.²⁶

Marc Tüngler, head of the German Association for the Protection of Securities Holders, laments the lack of political support necessary for innovation and economic restructuring. Germany is no longer internationally competitive in terms of electricity prices, for example. Politicians are responsible for this. Important levers would therefore be an improved location policy and a more innovation-friendly climate. We are far from the necessary solidarity between business and politics. He expects 2026 to be a year of decisions for politicians.²⁷

In his book „Wir Krisenakrobaten“ (We Crisis Acrobats), Stephan Grünewald, co-founder of the Cologne-based opinion research company Rheingold, describes the hope for self-efficacy that would enable our society to overcome the multitude of current crises. His recommendation consists of six points:

Truthfulness (clear identification of problems)
focus (successful national projects)
participation (making one’s own contribution clear)
fairness (unreasonable demands must be perceived as fair)
culture of debate (dealing more productively with changes in perspective) and
solidarity (which must be relearned).

Unfortunately, silo thinking („silodarity“) still prevails at present.²⁸

Martin Keller has returned to Germany from the US to become the new president of the Helmholtz Association. The association comprises 18 independent research centers with almost 48,000 employees and a budget of more than six billion euros. Keller wants to use a plan of action to ensure that Germany remains or becomes a global leader in selected fields of innovation. This requires closer cooperation, e.g., in the context of public-private partnerships (PPP), in which politics, research, and business cooperate in order to become more competitive. He believes it is time to break down old structures.²⁹

In his book „Visionen braucht das Land“ (The Country Needs Visions), Jochen Andritzky, co-initiator of Zukunft-Fabrik 2050, calls on politicians to develop visions of the future that can be discussed and provide guidance. This approach is more promising than short-term pseudo-solutions that merely combat the symptoms.³⁰ This return to the power of vision provides important impetus for management research, which in the past has often been content with incremental improvements. Design-oriented management research aims to be more practice-oriented in this regard.

Design-oriented management research in real-world laboratories of change

A research project at Würth has given rise to an AI start-up that could revolutionize the crafts. The aim of the research project conducted by the wholesaler of mounting and fastening material Würth and the AI Lab at the Technical University of Munich was to process inquiries from trade customers in sales more quickly. This led to the spin-off Mercura AI in March 2024, which uses AI to try to solve several problems:

Overcoming the shortage of skilled workers
increasing productivity for highly complex tasks, and
faster processing of inquiries and quotes.

Mercura AI combines semantic models, the recognition of requirements,
company-specific rules, and learning from previous quotes. The software processes both text and speech. The founders have combined AI expertise with industry experience. This example shows the potential of design-oriented management research in companies. ³¹

Nobel Prize winner Herbert Simon provided important impetus for design-oriented management research. His book „The Sciences of the Artificial,“ published in 1969, is not only a fundamental work on AI, but has also had a strong influence on design theory. The basic idea is that, in addition to the natural sciences, there is a universal science of design. This gave rise to the design methods movement. Not only the technical sciences, but also management science deal with the design of the possible (contingent). In the technical sciences, the design of new things is a natural goal. In management, political science, and social science, the diversity of individual systems and subsystems originating from humans has a specific complexity that is difficult to research purely empirically. Simon’s groundbreaking work emphasizes the interdisciplinarity of design.³²

Real-world laboratories of change open up new possibilities for management research. A real-world laboratory (living lab or sandbox) is a research and application space in practice where, for example, companies and their partners design innovative business models. In doing so, they combine research, learning, and action, promote interdisciplinary collaboration, and enable the testing of new legal frameworks (e.g., through the application of experimentation clauses). The concept became known in the 1990s primarily through the work of the Media Lab at the Massachusetts Institute of Technology (MIT).³³ In Europe, real-world laboratories are primarily intended to create modern forms of regulation (e.g., in urban development). Real-world laboratories have been used relatively little in management research to date. Empirical approaches dominate in dissertations. The advantage of real-world laboratories lies in their ability to better connect theory and practice.

Design-oriented management research is not only taking place at universities, but also increasingly in practice. University lecturers are increasingly supervising creative research approaches by employees in their companies. This approach is mainly used in bachelor’s and master’s theses in dual study programs, in which the course of study is organized in parallel with practical work. In the past, this has also been done more frequently in external dissertations and postdoctoral theses, e.g., by management consultants. The focus here was more on practical relevance. Solving complex problems requires research by interdisciplinary teams, whose members then receive their degrees in their respective fields. Universities should work with partners in the field to combine such projects into programs that can also build on each other (e.g., to design a sovereign AI from Europe).³⁴

The following figure summarizes various possible forms of design-oriented management research. Here, we distinguish between the type of degrees, the employment relationship of the researcher and the project and program types. In a part-time doctorate of a consulting employee, for example, it makes sense to compare the results of projects from several organizations and derive new insights from them. What seems important in this research approach is that design-oriented research projects based on theoretical foundations³⁵ now focus more strongly on concrete application in practice.

In 2026, we will further develop this approach to management research in the context of designing high-performance systems in which trustworthiness has become an important competitive advantage. One model for this is the start-up ecosystem in Munich, from which other regions can learn.³⁶

Conclusion

High-performance systems are characterized by their success and trustworthiness. In the current transition phase, Europe should seize this as an opportunity.
To do so, companies must master complex realignment processes and become more resilient, digital, and sustainable
Such connective strategic management (Strategy 5.0) is one of the fields of action of trustworthy high-performance systems
Another field of action is the promotion of high-performance teams with a growth mindset
This requires trustworthy partners and closer cooperation between business and politics
Real-world laboratories of change open up new opportunities for design-oriented management research.

Literature

[1] zu Fürstenberg, J., Kloepfer, I., How good we are is revealed in times of crisis – A wake-up call, Piper 2025

[2] Servatius, H.G., Trimming an organization for performance. In: Harvard Manager, 1988, No. 4, pp. 128-134

[3] Katzenbach, J.R., Smith, D.R., The wisdom of teams – Creating the high performance organization, Harvard Business School Press 1993

[4] de Waal, A., What makes a high performance organization, Warden Press 2019

[5] Servatius, H.G., AI as a tool for strategic management. In: Competivation Blog, May 1, 2025

[6] Sommer, U., US corporations are stronger than ever. In: Handelsblatt, December 29, 2025, pp. 1, 4-6

[7] Alvarez de Souza Soares, P., Holtermann, P., AMD wants to end Nvidia’s monopoly. In: Handelsblatt, January 7, 2026, pp. 18-19

[8] Knees, C., Disinformation as a business risk. In: Handelsblatt, January 7, 2026, pp. 20-21

[9] Merten, M., Companies sinking in AI junk. In: Handelsblatt, January 9, 2026, pp. 20-21

[10] Frankopan, P., „What does Europe have besides handbags and champagne?“ (Interview). In: Handelsblatt, December 19/20/21, 2025, pp. 12-13

[11] Fröndhoff, B., et al., Billions for restructuring. In: Handelsblatt, November 26, 2025, pp. 1, 4-5

[12] Servatius, H.G., Disruption of management education for AI-based realignments. In: Competivation Blog, October 10, 2025

[13] Servatius, H.G., Development and change in strategic management. In: Competivation Blog, September 19, 2025

[14] Servatius, H.G., Triple strategic realignment. In: Competivation Blog, June 7, 2024

[15] Servatius, H.G., Strategic leadership with contextual and relationship-oriented intelligence. In: Competivation Blog, March 14, 2023

[16] Huchzermeier, D. et al., Economy in reform gridlock. In: Handelsblatt, February 2/3/4, 2026, pp. 1, 6-7

[17] Servatius, H.G., Resilience-oriented strategic management. In: Competivation Blog, March 15, 2024

[18] Meiritz, A., „We will certainly remember a no.“ In: Handelsblatt, January 22, 2026, p. 1, 4-5

[19] Koch, M., Can an alliance of middle powers slow Trump down? In: Handelsblatt, January 22, 2026, p. 5

[20] Bomke, L., Germany’s AI hope. In: Handelsblatt, December 2, 2025, p. 1

[21] Czikszentmihalyi, M., Creativity – Flow and the psychology of discovery and invention, Harper Collins 1996

[22] Dweck, C., Mindset – The new psychology of success, Random House 2006

[23] Nadella, S., Hit Refresh – The quest to rediscover Microsoft’s soul and imagine a better future for everyone, Harper Collins 2017

[24] Murphy, M.C., Cultures of growth – How the new science of mindset can transform individuals, teams and organizations, Simon & Schuster 2024

[25] Herwig, M., Linnhoff, C., New A 45 bridge opened. In: Rheinische Post, December 23, 2025, p. A6

[26] Fukuyama, F., „Trumpism is a cry against modernity“ (interview). In: Handelsblatt, December 5/6/7, 2025, pp. 12-13

[27] Tüngler, M., „Friedrich Merz still has it in his hands“ (interview). In: Handelsblatt, December 11, 2025, pp. 22-23

[28] Grünewald, S., We crisis acrobats – Psychogram of an unsettled society, Kiepenheuer & Witsch 2025

[29] Delhaes, D., Architect of a German research breakthrough. In: Handelsblatt, December 30, 2025, p. 13

[30] Andritzky, J., The country needs visions – For long-term policies with the courage to face the future, Herder 2026

[31] Bomke, L., Revolutionizing the trade with AI. In: Handelsblatt, January 7, 2026, p. 26

[32] Simon, H.A., The sciences of the artificial, 3rd ed., MIT Press 1996

[33] Mitchell, W.J., City of bits – Space, place, and the infobahn, MIT Press 1995

[34] Servatius, H.G., AI and the future of management education. In: Competivation Blog, April 9, 2025

[35] Seckler, C., et al., Design sciences across industries – Building bridges for advancing impactful business research. In: Schmalenbach Journal of Business Research, December 9, 2025

[36] Banze, S., Freisinger, G.M., The Munich code. In: Manager Magazin, February 2026, pp. 30-36

AI as a tool for strategic management

01.05.25

Artificial intelligence (AI) is currently developing into a powerful tool for strategic management that accelerates, strengthens and changes learning processes. This applies to the corporate level as well as to the level of functional areas and business processes. Pioneering companies are using knowledge-specific AI in the various phases of strategic processes and achieving competitive advantages with innovative, AI-based business models. Generative AI has the character of a wake-up call.

In our series of blog posts on artificial intelligence, this article deals with the role of AI in strategic management. In it, I explain the increasing importance of AI in strategy processes.

Generative AI as a wake-up call

The use of artificial intelligence in strategic management is not new. Since the turn of the millennium, US digital companies such as Amazon have been using AI-based personalization as part of their innovative business models.¹ Surprisingly, many users of these business models are not aware of the contribution of AI.

In our book The Internet of Things and Artificial Intelligence as Game Changers, published in 2020, we described the strategy process for new IoT- and AI-based business models²and discussed relevant business model patterns.³At that time, however, interest in the topic was still limited in Germany.

The real wake-up call that shook the general public awake came in November 2022, when OpenAI released its ChatGPT dialog program. This action triggered a hype around generative AI and large language models, which was followed by a certain disillusionment.⁴

Many companies are now asking themselves what role artificial intelligence can play in their strategy processes.

AI-supported strategy processes at corporate level

A study by the Massachusetts Institute of Technology (MIT) concludes that artificial intelligence accelerates and strengthens learning processes.⁵ Such augmented learning builds on existing learning capabilities. An important field of application are the various phases of innovative strategy processes that help companies to gain a new form of competitive advantage.

It starts with an AI audit to analyze the company’s initial strategic situation and its use of AI. This is followed by AI-supported strategic foresight, which enables faster and more efficient early detection. Knowledge-based AI is also a means of realigning business models. Another phase is the design of an AI-oriented stakeholder ecosystem. When selecting partners, it is important to find the right balance between cooperation and competition.

Innovative AI platform architectures form the basis for relevant applications, and companies generally need partners to implement them. Strategies are implemented with the help of agile, AI-supported performance management. This involves close coordination between the corporate level and the level of connected business processes.

Strategic learning loops, which take the form of rapid iterations, play a decisive role in agile strategy processes. This turns the analysis of the initial strategic situation into a dynamic process.

AI audit to analyze the initial strategic situation

A study by the German Economic Institute (IW) concludes that AI could contribute 330 billion euros to gross value added nationwide. One in five companies already uses AI. However, most applications are rather selective, e.g. in the form of chatbots for customer inquiries. Surprisingly, 66% of companies say that AI is not relevant to their business model. 36 percent consider integration into existing systems to be difficult. 47% complain about the lack of employee expertise. NRW Minister President Hendrik Wüst nevertheless believes that AI could be the driving force behind an economic upturn.⁶

To achieve this goal, companies should carry out an AI audit and use a SWOT analysis, for example, to gain an overview of their initial strategic situation.⁷ Interestingly, results of such an analysis of strengths, weaknesses, opportunities and threats are similar. One strength of companies is that they have a lot of specific knowledge that has the potential to be enhanced by AI. This is often offset by weaknesses in the systematic anchoring of AI in strategies and processes. The potential of AI lies both in increasing productivity and in innovation benefits through new products, services and business models. On the other hand, there are many threats from competitors, foreign stakeholder ecosystems and misuse of the power inherent in artificial intelligence.⁸

On this basis, the next step is to prepare even better for future developments with the help of AI-supported strategic foresight.

AI-supported strategic foresight

The term strategic foresight, coined in the 1980s, has a long history, during which methods such as scenario analysis, which are still widely used today, were developed. The Gamechanger Radar developed by us makes it possible to prepare for far-reaching changes.⁹With AI-supported strategic foresight, pioneering companies are now writing a new chapter in foresight. This chapter assumes a change in the way people search for information on the internet.

For example, Google has developed the new search function „Overview with AI“, which provides summarized texts on topics. An example is shown in the following illustration. The topic I entered is: „Applying Complexity Theory in Management“. The answer that Google provides is surprisingly good. It describes the paradigm shift in strategic management that has taken place in recent decades more comprehensively and better than many individual publications on this topic.

Foresight users will learn to improve their prompting capabilities relatively quickly. In addition, AI-supported foresight platforms are currently emerging that simplify and accelerate the early recognition of new trends, which usually take the form of weak signals.

Of course, this development also poses a threat to Google’s traditional search engine business, which is linked to advertising. The start-up Perplexity, for example, is trying to take users away from Google with its user-friendly „answer engine“. It remains to be seen what effect this will have on the market leader’s profit driver¹⁰

Reasoning AI enables advantages for complex tasks such as strategic foresight. It is now offered by some AI developers. In reasoning, the AI breaks down possible queries into sub-problems and processes them step by step. Such slower thinking costs more computing power and electricity. Developers call the „reasoning“ of AI a chain of thought (CoT). Reasoning models achieve this through an additional training step that uses reinforcement learning to train detailed reasoning. Similar to an experienced employee, reasoning models analyze complex information step by step. To do this, they need a single precise prompt and a lot of context. However, the application of reasoning AI in strategic foresight is still at the experimental stage.¹¹

AI-based realignment of business models

Innovative business models for AI-based robotics are currently emerging. This represents an opportunity for Europe. Stanford professor and great „godmother of AI“ Fei-Fei Li has founded the start-up World Labs, which develops AI models for the spatial intelligence of robots that support machines. Google subsidiary DeepMind and digital giant Nvidia are also working on partner networks for AI-based human-like robots. Many of the partners come from Europe. In addition to well-known robotics companies, start-ups such as Anybotics (Switzerland) and Agile Robots, Neura Robotics and Quantum Systems from Germany are emerging here, although they do not have as much funding as their competitors from the USA (e.g. Figure AI and Covariant). For Europe, it is important to seize the opportunities arising from the combination of in-depth industry-specific knowledge and innovative AI models as quickly as possible.¹²

Two dimensions are relevant for an AI-based realignment of business models. These dimensions are productivity orientation and innovation orientation. Most companies start with an AI-based increase in productivity and use AI in routine processes to reduce personnel costs. In addition, many fields of application for AI-based innovations have now emerged. When both dimensions come together, we speak of AI-based ambidexterity. The term ambidexterity originally refers to the ability to use both hands in sport. Applied to management, ambidextrous leadership describes leadership that strikes a good balance between innovation and productivity.¹³

The specific applications of these two dimensions in industries and companies result in a wide variety of AI-based ambidexterity. The new business models are embedded in AI-oriented stakeholder ecosystems.

AI-oriented stakeholder ecosystems

German and European policymakers are planning to boost the performance of their AI ecosystem. In view of the changing geopolitical situation, the coalition agreement of the new German government provides for a strengthening of digital sovereignty. The digital policy of the European Union (EU) aims in the same direction. Five gigantic data centers are planned in order to catch up in the field of artificial intelligence. The Jülich and Stuttgart sites are candidates for such a gigafactory in Germany. When it comes to AI regulation, the EU wants to focus more on competitiveness and reduce bureaucracy. An EU action plan has been drafted to this end. It remains to be seen whether these measures will be enough to reduce dependence on the large cloud providers (hyperscalers) from the USA.¹⁴

There are also two dimensions to consider when designing a company’s AI-oriented stakeholder ecosystem.¹⁵ One dimension is the dependence on powerful AI providers. In order to reduce this dependency, the second dimension for companies is improving their own skills in the development and application of artificial intelligence. In the hype phase of basic AI models, dependence on US providers has increased. The opportunity for Europe now lies primarily in knowledge-specific AI models for various applications. Hybrid AI ecosystems are emerging by connecting these two dimensions. Such connectivity requires specific skills.

In view of the geopolitical uncertainties, companies are faced with the difficult task of finding the right partners when designing their AI ecosystem. The transitions between cooperation and competition are fluid. The term coopetition describes such a situation.¹⁶However, the theoretical basis for a combination of cooperation and competition is still lacking in AI ecosystems. An important field of application is the selection and in-house development of innovative AI platform architectures.

Innovative AI platform architectures

The chip manufacturer AMD and the Finnish start-up Silo AI, which belongs to AMD, are working together with the companies of the Swedish Wallenberg Group. The Nvidia competitor AMD has announced a partnership with 38 companies. These include AstraZeneca, Scania, Saab, Ericsson and IKEA. The collaboration is coordinated by the Wallenberg innovation network Combient. The aim is to scale company-specific AI models. While OpenAI trains its AI models on Nvidia chips, Silo AI uses chips from AMD. The role of Silo AI is to accelerate the deployment of AI models at companies that use AMD platforms. The infrastructure on which the work has begun plays an important role here, as a move is time-consuming. Silo AI uses multimodal AI agents, i.e. models that process images and audio files as well as speech.¹⁷

Established digital companies have been practising an organizational form with an IT platform at its center for some time now.¹⁸With the increasing importance of artificial intelligence, this concept is becoming more and more relevant for established companies. Innovative AI platform architectures combine both the strategic and operational levels as well as centralized and decentralized organizational units. This enables all business processes and projects to have access to a common database. Due to their connecting role, AI platforms not only become a strategic building block, but also an important organizational design element. One question that is not easy to answer is how large the share of partners and the company’s own share should be in such an AI platform.

Innovative platform architectures also provide the infrastructure for AI-supported performance management.

AI-supported performance management

To answer the question of how artificial intelligence can improve performance management, it helps to take a look at the history of performance measurement. The Management by Objectives (MbO) developed by Peter Drucker and the goal-setting theory developed by organizational psychologist Edwin Locke provide important conceptual foundations. Back in the 1980s, Intel developed the agile Objectives and Key Results (OKR) method, which the venture capitalist Kleiner Perkins used at Google, for example.¹⁹In Germany, the Balanced Scorecard method, which emerged from a best practice study by Robert Kaplan and David Norton, is much better known.²⁰An AI-supported performance management system designed by Kleiner Perkins and the start-up Betterworks now aims to better connect strategy and motivation.

Although artificial intelligence is one of the top management issues for 2025, many companies neither formulate specific AI targets nor measure the results. A global BCG study, in which 1,800 managers were surveyed, found that only 24% of companies track their operational and financial AI targets. AI-supported performance management faces three challenges. These challenges are:²¹

Do not stall early trials
define appropriate key results for the success of an individual measure and, in addition
capture the longer-term effects resulting from the interaction of various measures.

The agile OKR method provides a conceptual basis for this, but requires adaptation. OKR pioneer Kleiner Perkins is one of the investors in performance management software provider Betterworks. The vision of the Palo Alto-based company, which was founded in 2013, is to further develop traditional performance management. AI plays an important role here as a co-pilot. Managers can thus invest time saved on routine tasks in better harmonization of strategic and operational projects. Important use cases are:²²

Alignment of ambitious corporate goals and personal goals
data-based, motivating feedback and
the support of communication and learning processes.

The intended benefit, which contributes to the overall success, is

a reduction in bias, more fairness and objectivity
increased productivity and
better personal relationships.

This brings performance management one step closer to the motivational concept already pursued by goal-setting theory.

With the increasing importance of artificial intelligence in strategic management, geopolitical expertise in working with stakeholders is becoming ever more important alongside practical skills in using AI as a tool. One basis for this is a strong future narrative.

A strong future narrative as a basis

In our 2020 book on the gamechanging potential of artificial intelligence, we took a critical look at European and German digital policy.²³The new German government now faces the task of developing a strong future narrative that connects various policy areas.²⁴One approach to such a much-needed grand narrative is the application of trustworthy AI both to increase productivity and to solve the innovation and environmental problems of organizations. At the heart of this is the new form of ambidexterity outlined earlier.

Traditional ambidexterity strives for a balance between tapping innovation potential (exploration) and utilization of productivity (exploitation). With the help of AI, which should be trustworthy, it is now possible to simultaneously

reduce labor costs by increasing productivity, counter the shortage of skilled workers²⁵and
to make greater use of qualified personnel for the digital and ecological realignment of organizations²⁶

In view of the changed geopolitical situation, there is a window of opportunity for AI made in Europe, which the „old continent“ should use to strive for global market leadership in the necessary sustainability innovations.²⁷Due to the large number of crises to be overcome, this initially requires resilience-oriented strategic management.²⁸

Conclusion

Strategy processes become more efficient through the use of artificial intelligence
Knowledge-specific AI supports strategic foresight, the realignment of business models, the design of stakeholder ecosystems, innovative platform architectures and performance management
Pioneering companies are working on AI-based ambidextry
In view of the geopolitical challenges, choosing the right partners is crucial.

Literature

[1] Servatius, H.G., Competitive advantages with knowledge-specific AI. In: Competivation Blog, 11.02.2025

[2] Kaufmann, T., Servatius, H.G., Das Internet der Dinge und Künstliche Intelligenz als Game Changer – Wege zu einem Management 4.0 und einer digitalen Architektur, SpringerVieweg 2020, p. 56ff.

[3] Kaufmann, Servatius, op. cit. p. 34ff.

[4] Servatius, H.G., Development of AI technologies. In: Competivation Blog, 19.02.2025

[5] Alavi, M., Westerman, G., How GenAI Will Transform Knowledge Work. In: Harvard Business Review, November 7, 2023

[6] Höning, A., Kowalewski, R., Every fifth company in NRW uses AI. In: Rheinische Post, November 13, 2025, p. 1

[7] Servatius, H.G., Auditing the innovation system of a company. In: Competivation Blog, 19.03.2015

[8] Suleyman, M., Bhaskar, M., The Coming Wave – Technology, Power and the Twenty-First Century’s Greatest Dilemma, Crown 2013

[9] Servatius, H.G., Strategic foresight with a game changer radar. In: Competivation Blog, 27.01.2021

[10] Alvares de Souza Soares, P., Geldmaschine Google – Wie lange noch? In: Handelsblatt, April 25/26/27, 2025, p. 26-27

[11] Knees, L., Why users pay more for slow AI. In: Handelsblatt, March 31, 2025, pp. 24-25

[12] Holtermann, F., Schimroszik, N., The robots are coming! In: Handelsblatt, January 3/4/5, 2025, pp. 44-48

[13] O’Reilley, C., Tushman, M., Lead and Disrupt – How to Solve the Innovator’s Dilemma, Stanford Business Books 2016

[14] Bomke, L., et al, Europe wants to build its own AI factories. In: Handelsblatt, April 9, 2025, p. 6-7

[15] Servatius, H.G., Designing innovative stakeholder ecosystems. In: Competivation Blog, 10.01.2023

[16] Brandenburger, A.M., Nalebuff, B.J., Co-Opetition – A Revolutionary Mindset That Combines Competition and Co-Operation, Bantam 1996

[17] Holzki, L., AMD enters into partnership with the industry. In: Handelsblatt, January 30, 2025, p. 24

[18] Servatius, H.G., The resource platform with agile teams as a new organizational form. In: Competivation Blog, 12.01.2021

[19] Doerr, J., Measure What Matters – How Google, Bono and the Gates Foundation Rock the World with OKRs, Portfolio/Penguin 2018

[20] Kaplan, R.S., Norton, D.P., Balanced Scorecard – Translating Strategy into Action, Harvard Business School Press 1996

[21] Bomke, L., Höppner, A., Only a few companies measure their AI initiatives. In: Handelsblatt, January 16, 2025, p. 21

[22] Gouldsberry, M., The Pivotal Role of AI in Performance Management, January 11, 2025

[23] Kaufmann, Servatius, op. cit. p. 203ff.

[24] Servatius, H.G., On the way to a new economic policy narrative. In: Competivation Blog, 16.05.2022

[25] Servatius, H.G., Process-oriented AI to increase productivity. In: Competivation Blog, 12.03.2025

[26] Servatius, H.G., AI and the future of management education. In: Competivation Blog, 09.04.2025

[27] Servatius, H.G., Sustainability-oriented strategic management. In: Competivation Blog, 15.08.2024

[28] Servatius, H.G., Resilience-oriented strategic management. In: Competivation Blog, 15.03.2024

KI als Werkzeug für das strategische Management

01.05.25

Die Künstliche Intelligenz (KI) entwickelt sich gegenwärtig zu einem mächtigen Werkzeug für das strategisches Management, das Lernprozesse beschleunigt, verstärkt und verändert. Dies gilt sowohl für die Unternehmensebene als auch für die Ebene der Funktionsbereiche und Geschäftsprozesse. Vorreiter-Unternehmen setzen eine wissensspezifische KI in den verschiedenen Phasen von Strategieprozessen ein und erzielen Wettbewerbsvorteile mit innovativen, KI-basierten Geschäftsmodellen. Dabei hat die generative KI den Charakter eines Weckrufs.

In unserer Blogpost-Reihe zur Künstlichen Intelligenz beschäftigt sich dieser Beitrag mit der Rolle von KI im strategischen Management. Darin erläutere ich die zunehmende Bedeutung von KI in Strategieprozessen.

Generative KI als Weckruf

Die Anwendung von Künstlicher Intelligenz im strategischen Management ist nicht neu. Bereits seit der Jahrtausendwende haben US-amerikanische Digital-Unternehmen wie Amazon die KI-basierte Personalisierung im Rahmen ihrer innovativen Geschäftsmodelle eingesetzt.¹ Erstaunlicherweise ist der Beitrag der KI vielen Nutzern dieser Geschäftsmodelle nicht bewusst.

In unserem 2020 erschienenen Buch Das Internet der Dinge und Künstliche Intelligenz als Game Changer haben wir den Strategieprozess für neue IoT- und KI-basierte Geschäftsmodelle beschrieben²und relevante Geschäftsmodellmuster behandelt.³Zu dieser Zeit hielt sich das Interesse an dem Thema in Deutschland allerdings noch in Grenzen.

Der eigentliche Weckruf, der dann eine breite Öffentlichkeit wachgerüttelt hat, ist im November 2022 erfolgt, als OpenAI sein Dialogprogramm ChatGPT veröffentlichte. Diese Aktion löste einen Hype um die generative KI und große Sprachmodelle aus, dem eine gewisse Ernüchterung gefolgt ist.⁴

Viele Unternehmen fragen sich nun, welche Rolle die Künstliche Intelligenz in ihren Strategieprozessen spielen kann.

KI-unterstützte Strategieprozesse auf der Unternehmensebene

Eine Studie des Massachusetts Institute of Technology (MIT) kommt zu dem Ergebnis, dass Künstliche Intelligenz Lernprozesse beschleunigt und verstärkt.⁵ Ein solches erweitertes (augmented) Lernen setzt an den vorhandenen Lernfähigkeiten an. Ein wichtiges Anwendungsfeld sind die verschiedenen Phasen von innovativen Strategieprozessen, die Unternehmen zu einer neuen Form von Wettbewerbsvorteilen verhelfen.

Am Anfang steht ein KI-Audit zur Analyse der strategischen Ausgangssituationen des Unternehmens und seiner Anwendung von KI. Hieran schließt sich eine KI-unterstützte strategische Vorausschau (Foresight) an, die eine schnellere und leistungsfähigere Früherkennung ermöglicht. Die wissensspezifische KI ist auch ein Mittel bei der Neuausrichtung von Geschäftsmodellen. Eine weitere Phase ist die Gestaltung eines KI-orientierten Stakeholder-Ökosystems. Bei der Auswahl von Partnern gilt es, die richtige Balance zwischen Kooperation und Wettbewerb zu finden.

Eine Basis für relevante Anwendungen bilden innovative KI-Plattform-Architekturen, zu deren Realisierung Unternehmen in der Regel Partner benötigen. Die Umsetzung von Strategien erfolgt mit Hilfe eines agilen, KI-unterstützten Performance Managements. Dabei findet eine enge Abstimmung zwischen der Unternehmensebene und der Ebene verbundener Geschäftsprozesse statt.

Eine entscheidende Rolle bei agilen Strategieprozessen spielen strategische Lernschleifen, die in Form von schnellen Iterationen ablaufen. So wird die Analyse der strategischen Ausgangssituation zu einem dynamischen Prozess.

KI-Audit zur Analyse der strategischen Ausgangssituation

Eine Studie des Instituts der deutschen Wirtschaft (IW) kommt zu dem Ergebnis, KI könne bundesweit 330 Milliarden Euro zur Bruttowertschöpfung beitragen. Jedes fünfte Unternehmen setzt bereits KI ein. Die meisten Anwendungen sind aber eher punktuell, z.B. in Form von Chatbots für Kundenanfragen. Erstaunlicherweise sagen 66 Prozent der Unternehmen, KI sei für ihr Geschäftsmodell nicht relevant. 36 Prozent halten die Integration in bestehende Systeme für schwierig. Über das fehlende Know-how der Beschäftigten klagen 47 Prozent. Der NRW-Ministerpräsident Hendrik Wüst glaubt aber dennoch, KI könne der Motor für einen wirtschaftlichen Aufschwung sein.⁶

Um dieses Ziel zu erreichen, sollten Unternehmen ein KI-Audit durchführen und sich z.B. mit der SWOT-Analyse einen Überblick zu ihrer strategischen Ausgangssituation verschaffen.⁷ Interessanterweise ähneln sich die Ergebnisse einer solchen Analyse der Stärken, Schwächen, Möglichkeiten und Bedrohungen. Eine Stärke vieler Unternehmen ist, dass sie über viel spezifisches Wissen verfügen, welches das Potenzial zu einer Erweiterung durch KI hat. Dem stehen häufig Schwächen bei einer systematischen Verankerung von KI in Strategien und Prozessen gegenüber. Die Möglichkeiten von KI liegen sowohl in der Produktivitätssteigerung als auch in Innovationsvorteilen durch neue Produkte, Dienstleistungen und Geschäftsmodelle. Andererseits gibt es vielfältige Bedrohungen durch Konkurrenten, nicht-europäische Stakeholder-Ökosysteme und einen Missbrauch der in Künstlicher Intelligenz steckenden Macht.⁸

Auf dieser Grundlage geht es dann in einem nächsten Schritt darum, sich mit Hilfe einer KI-unterstützten strategischen Vorausschau noch besser auf zukünftige Entwicklungen vorzubereiten.

KI-unterstützte strategische Vorausschau

Die in den 1970er und 80er Jahren geprägten Begriffe strategische Früherkennung und Vorausschau (Foresight) haben eine längere Vorgeschichte, in der noch heute verbreitete Methoden wie die Szenarioanalyse entstanden sind. Der von uns entwickelte Gamechanger-Radar ermöglicht eine Vorbereitung auf tiefgreifende Veränderungen.⁹ Mit einer KI-unterstützten strategischen Vorausschau schreiben Vorreiter-Unternehmen nun ein neues Foresight-Kapitel. Dieses Kapitel geht von einem Wandel der Art und Weise aus, wie Menschen im Internet nach Informationen suchen.

So hat Google die neue Suchfunktion „Übersicht mit KI“ entwickelt, die zusammenfassende Texte zu Themen liefert. Ein Beispiel ist in der folgenden Abbildung dargestellt. Das Thema, das ich eingegeben habe, lautet: „Applying Complexity Theory in Management“. Die Antwort, die Google liefert, ist überraschend gut. Sie beschreibt den Paradigmenwechsel im strategischen Management, der sich in den vergangenen Jahrzehnten vollzogen hat, umfassender und besser als viele einzelne Publikationen zu diesem Thema.

Foresight-Anwender werden relativ schnell lernen, ihre Prompting-Fähigkeiten zu verbessern. Daneben entstehen gegenwärtig KI-unterstützte Foresight-Plattformen, die das frühzeitige Erkennen neuer Trends, die sich meist in Form von schwachen Signalen ankündigen, vereinfachen und beschleunigen.

Natürlich stellt diese Entwicklung auch eine Bedrohung für das traditionelle, mit Werbung verknüpfte Suchmaschinengeschäft von Google dar. Das Start-up Perplexity versucht z.B. mit seiner benutzerfreundlichen „Antwortmaschine“, Google Nutzer abzujagen. Es bleibt abzuwarten, wie sich dies auf den Gewinnbringer des Marktführers auswirken wird.¹⁰

Für komplexe Aufgabenstellungen wie die strategische Vorausschau bietet die Reasoning AI („argumentierende KI“) Vorteile. Sie wird inzwischen von einigen KI-Entwicklern angeboten. Beim Reasoning zerlegt die KI mögliche Anfragen in Teilprobleme und bearbeitet diese schrittweise. Ein solches langsameres Denken kostet mehr Computerleistung und Strom. Das „Nachdenken“ von KI nennen Entwickler Chain of Thought (CoT) im Sinne einer Argumentationskette. Reasoning-Modelle erreichen dies durch einen zusätzlichen Trainingsschritt, der mit Hilfe des Reinforcement Learning ausführliche Begründungen schult. Ähnlich wie ein erfahrener Mitarbeiter analysieren Reasoning-Modelle schrittweise komplexe Informationen. Dazu benötigen sie einen einzigen präzisen Prompt und viel Kontext. Die Anwendung von „argumentierender KI“ bei der strategischen Vorausschau befindet sich allerdings noch im Experimentierstadium.¹¹

KI-basierte Neuausrichtung von Geschäftsmodellen

Gegenwärtig entstehen innovative Geschäftsmodelle für eine KI-basierte Robotik. Hierin liegt eine Chance für Europa. Die Stanford-Professorin und große „Patin der KI“ Fei-Fei Li hat das Start-up World Labs gegründet, das KI-Modelle für eine räumliche Intelligenz von Robotern entwickelt, die Maschinen unterstützen. Auch die Google-Tochter DeepMind und der Digital-Gigant Nvidia arbeiten an Partnernetzwerken für KI-basierte menschenähnliche Roboter. Viele der Partner kommen aus Europa. Neben bekannten Robotik-Unternehmen entstehen hier Start-ups wie Anybotics (Schweiz) sowie Agile Robots, Neura Robotics und Quantum Systems aus Deutschland, die aber nicht über so große finanzielle Mittel verfügen, wie ihre Wettbewerber aus den USA (z.B. Figure AI und Covariant). Für Europa kommt es darauf an, möglichst schnell die Chancen zu nutzen, die sich aus der Verbindung von tiefem branchenspezifischem Wissen und innovativen KI-Modellen ergeben.¹²

Bei einer KI-basierten Neuausrichtung von Geschäftsmodellen sind zwei Dimensionen relevant. Diese Dimensionen sind die Produktivitätsorientierung und die Innovationsorientierung. Die meisten Unternehmen beginnen mit einer KI-basierten Produktivitätssteigerung und setzen KI bei Routineprozessen ein, um Personalkosten zu senken. Daneben sind inzwischen viele Anwendungsfelder für KI-basierte Innovationen entstanden. Wenn beide Dimensionen zusammenkommen, sprechen wir von einer KI-basierten Ambidextrie. Der Begriff Ambidextrie kennzeichnet ursprünglich im Sport die Fähigkeit zum Einsatz beider Hände. Übertragen auf das Management beschreibt Ambidextrous Leadership eine Führung, die eine gute Balance zwischen Innovation und Produktivität findet.¹³

Aufgrund der spezifischen Anwendungen dieser beiden Dimensionen in Branchen und Unternehmen ergibt sich eine große Vielfalt an KI-basierter Ambidextrie. Dabei sind die neuen Geschäftsmodelle in KI-orientierte Stakeholder-Ökosysteme eingebettet.

KI-orientierte Stakeholder-Ökosysteme

Die deutsche und die europäische Politik planen eine Leistungssteigerung ihres KI-Ökosystems. Angesichts einer sich wandelnden geopolitischen Lage sieht der Koalitionsvertrag der neuen Bundesregierung eine Stärkung der digitalen Souveränität vor. Die Digitalpolitik der Europäischen Union (EU) zielt in die gleiche Richtung. Geplant sind fünf riesige Rechenzentren, um den Rückstand bei der Künstlichen Intelligenz aufzuholen. Kandidaten für eine solche Gigafactory in Deutschland sind die Standorte Jülich und Stuttgart. Bei der KI-Regulierung möchte die EU die Wettbewerbsfähigkeit stärker in den Mittelpunkt stellen und Bürokratie abbauen. Hierzu liegt der Entwurf eines EU-Aktionsplans vor. Ob diese Maßnahmen ausreichen, um die Abhängigkeit von den großen Cloud-Anbietern (Hyperscaler) aus den USA zu verringern, bleibt abzuwarten.¹⁴

Auch bei einer Gestaltung des KI-orientierten Stakeholder-Ökosystems eines Unternehmens¹⁵ sind zwei Dimensionen zu beachten. Die eine Dimension ist die Abhängigkeit von mächtigen, nicht-europäischen KI-Anbietern. Um diese Abhängigkeit zu verringern, gewinnt als zweite Dimension für Unternehmen eine Verbesserung der eigenen Kompetenzen zur Entwicklung und Anwendung von Künstlicher Intelligenz an Bedeutung. In der Hype-Phase von KI-Grundlagenmodellen hat die Abhängigkeit von US-amerikanischen Anbietern zugenommen. Die Chance für Europa liegt nun vor allem bei wissensspezifischen KI-Modellen für verschiedene Anwendungen. Durch eine Verbindung dieser beiden Dimensionen entstehen hybride KI-Ökosysteme. Eine solche Konnektivität erfordert spezifische Fähigkeiten.

Angesichts der geopolitischen Unsicherheiten stehen Unternehmen bei der Gestaltung ihres KI-Ökosystems vor der schweren Aufgabe, die richtigen Partnern zu finden. Dabei sind die Übergänge zwischen Kooperation und Wettbewerb fließend. Eine solche Situation beschreibt der Begriff Coopetition.¹⁶ Für eine Kombination von Cooperation und Competition fehlen bei KI-Ökosystemen bislang aber noch die theoretischen Grundlagen. Ein wichtiges Anwendungsfeld ist die Auswahl und eigene Entwicklung von innovativen KI-Plattform-Architekturen.

Innovative KI-Plattform-Architekturen

Der Chiphersteller AMD und das zu AMD gehörende finnische Start-up Silo AI arbeiten mit den Unternehmen der schwedischen Wallenberg-Gruppe zusammen. Der Nvidia-Wettbewerber AMD hat eine Partnerschaft mit 38 Unternehmen bekannt gegeben. Hierzu gehören u.a. AstraZeneca, Scania, Saab, Ericsson und IKEA. Die Zusammenarbeit koordiniert das Wallenberg-Innovationsnetzwerk Combient. Das Ziel ist eine Skalierung unternehmensspezifischer KI-Modelle. Während OpenAI seine KI-Modelle auf Nvidia-Chips trainiert, verwendet Silo AI Chips von AMD. Die Rolle von Silo AI ist, den Einsatz von KI-Modellen bei Unternehmen, die AMD-Plattformen nutzen, zu beschleunigen. Eine wichtige Rolle spielt dabei, auf welcher Infrastruktur die Arbeiten begonnen haben, da ein Umzug aufwändig ist. Silo AI setzt multimodale KI-Agenten ein, also Modelle, die neben Sprache auch Bilder und Audiodateien verarbeiten.¹⁷

Etablierte Digital-Unternehmen praktizieren seit geraumer Zeit eine Organisationsform, in deren Zentrum sich eine IT-Plattform befindet.¹⁸ Mit der zunehmenden Bedeutung von Künstlicher Intelligenz wird dieses Konzept für etablierte Unternehmen immer relevanter. Dabei verbinden innovative KI-Plattform-Architekturen sowohl die strategische und die operative Ebene als auch zentrale und dezentrale Organisationseinheiten. Dies ermöglicht, dass alle Geschäftsprozesse und Projekte Zugang zu einer gemeinsamen Datenbasis haben. Aufgrund ihrer verbindenden Rolle werden KI-Plattformen somit nicht nur zu einem Strategiebaustein, sondern auch zu einem wichtigen organisatorischen Gestaltungselement. Eine nicht einfach zu beantwortende Frage ist, wie groß der Anteil von Partnern und der eigene Anteil bei einer solchen KI-Plattform sein soll.

Innovative Plattform-Architekturen liefern auch die Infrastruktur für ein KI-unterstütztes Performance Management.

KI-unterstütztes Performance Management

Bei der Beantwortung der Frage, wie Künstliche Intelligenz das Performance Management verbessern kann, hilft ein Blick in die Geschichte der Leistungsmessung. Wichtige konzeptionelle Grundlagen liefern das von Peter Drucker entwickelte Management bei Objectives (MbO) und die von dem Organisationspsychologen Edwin Locke stammende Zielsetzungstheorie. Bereits in den 1980er Jahren entstand bei Intel die agile Objectives and Key Results (OKR-) Methode, die der Wagniskapitalgeber Kleiner Perkins z.B. bei Google einsetzte.¹⁹ In Deutschland wesentlich bekannter ist die aus einer Best-Practice-Studie von Robert Kaplan und David Norton hervorgegangene Balanced-Scorecard-Methode.²⁰ Ein von Kleiner Perkins und dem Start-up Betterworks gestaltetes, KI-unterstütztes Performance Management zielt nun auf eine bessere Verbindung von Strategie und Motivation ab.

Für das Jahr 2025 gehört die Künstliche Intelligenz zwar zu den Topthemen des Managements, viele Unternehmen formulieren aber weder konkrete KI-Ziele noch messen sie die Ergebnisse. Eine weltweite BCG-Studie, in der 1800 Manager befragt wurden, kommt zu dem Ergebnis, dass nur 24 Prozent der Unternehmen ihre operativen und finanziellen KI-Ziele nachverfolgen. Ein KI-unterstütztes Performance Management steht vor drei Herausforderungen. Diese Herausforderungen sind:²¹

Frühe Erprobungsversuche nicht abwürgen
geeignete Schlüsselergebnisse für den Erfolg einer einzelnen Maßnahme festlegen und darüber hinaus
die längerfristigen Effekte erfassen, die aus dem Zusammenwirken verschiedener Maßnahmen resultieren.

Die agile OKR-Methode liefert hierfür eine konzeptionelle Basis, bedarf aber einer Anpassung. Der OKR-Pionier Kleiner Perkins ist einer der Kapitalgeber des Anbieters von Performance-Management-Software Betterworks. Die Vision des 2013 gegründeten Unternehmens mit Sitz in Palo Alto ist, das traditionelle Performance Management weiterzuentwickeln. Eine wichtige Rolle spielt dabei KI als Co-Pilot. Zeitgewinne bei Routineaufgaben können Manager so in eine bessere Harmonisierung strategischer und operativer Projekte investieren. Wichtige Anwendungsfälle (Use Cases) sind:²²

Eine Abstimmung von anspruchsvollen Unternehmenszielen und persönlichen Zielen
datenbasierte, motivierende Feedbacks sowie
die Unterstützung von Kommunikations- und Lernprozessen.

Der angestrebte Nutzen, der zum Gesamterfolg beiträgt, ist:

Eine Verringerung von Voreingenommenheit (bias), mehr Fairness und Objektivität
eine erhöhte Produktivität sowie
bessere persönliche Beziehungen.

Damit kommt das Performance Management dem schon von der Zielsetzungstheorie verfolgten Motivationsgedanken einen Schritt näher.

Mit der zunehmenden Bedeutung von Künstlicher Intelligenz im strategischen Management wird neben praktischen Fähigkeiten beim Umgang mit KI als Werkzeug die geopolitische Kompetenz bei der Zusammenarbeit mit Stakeholdern immer wichtiger. Eine Grundlage hierfür ist ein starkes Zukunftsnarrativ.

Ein starkes Zukunftsnarrativ als Grundlage

In unserem 2020 erschienenen Buch zum Gamechanger-Potenzial der Künstlichen Intelligenz haben wir uns kritisch mit der europäischen und der deutschen Digitalpolitik auseinandergesetzt.²³ Die neue Bundesregierung steht nun vor der Aufgabe, ein starkes Zukunftsnarrativ zu entwickeln, das verschiedene Politikfelder verbindet.²⁴ Ein Ansatz zu einer solchen dringend benötigten, großen Erzählung ist die Anwendung von vertrauenswürdiger KI sowohl zur Steigerung der Produktivität als auch zur Lösung der Innovations- und Umweltprobleme von Organisationen. Im Mittelpunkt steht dabei die bereits skizzierte neue Form der Ambidextrie.

Die traditionelle Ambidextrie strebt eine Balance zwischen der Erschließung von Innovationspotenzialen (Exploration) und der Ausschöpfung von Produktivität (Exploitation) an. Mit Hilfe einer KI, die vertrauenswürdig sein sollte, bietet sich nun die Möglichkeit, gleichzeitig

durch Produktivitätssteigerungen die Arbeitskosten zu senken, dem Fachkräftemangel zu begegnen²⁵ und
qualifiziertes Personal stärker zur digitalen und ökologischen Neuausrichtung von Organisationen einzusetzen.²⁶

Angesichts der veränderten geopolitischen Lage ergibt sich für eine AI made in Europe ein Zeitfenster, das der „alte Kontinent“ nutzen sollte, um die Weltmarktführerschaft bei notwendigen Nachhaltigkeitsinnovationen anzustreben.²⁷ Aufgrund der Vielzahl der zu bewältigenden Krisen erfordert dies zunächst ein resilienzorientiertes strategisches Management.²⁸

Fazit

Strategieprozesse werden durch die Anwendung von Künstlicher Intelligenz leistungsfähiger
Eine wissensspezifische KI unterstützt die strategische Vorausschau, eine Neuausrichtung von Geschäftsmodellen, die Gestaltung von Stakeholder-Ökosystemen, innovative Plattform-Architekturen und das Performance Management
Vorreiter-Unternehmen arbeiten an einer KI-basierten Ambidextrie
Angesichts der geopolitischen Herausforderungen kommt es entscheidend auf die Wahl der richtigen Partner an.

Literatur

[1] Servatius, H.G., Wettbewerbsvorteile mit wissensspezifischer KI. In: Competivation Blog, 11.02.2025

[2] Kaufmann, T., Servatius, H.G., Das Internet der Dinge und Künstliche Intelligenz als Game Changer – Wege zu einem Management 4.0 und einer digitalen Architektur, SpringerVieweg 2020, S. 56ff.

[3] Kaufmann, Servatius, a.a.O., S. 34ff.

[4] Servatius, H.G., Entwicklung der KI-Technologien. In: Competivation Blog, 19.02.2025

[5] Alavi, M., Westerman, G., How GenAI Will Transform Knowledge Work. In: Harvard Business Review, 7. November 2023

[6] Höning, A., Kowalewski, R., Jeder fünfte Betrieb in NRW nutzt KI. In: Rheinische Post, 13. November 2025, S. 1

[7] Servatius, H.G., Auditierung des Innovationssystems eines Unternehmens. In: Competivation Blog, 19.03.2015

[8] Suleyman, M., Bhaskar, M., The Coming Wave – Technology, Power and the Twenty-First Century‘s Greatest Dilemma, Crown 2023

[9] Servatius, H.G., Strategische Vorausschau mit einem Game-Changer-Radar. In: Competivation Blog, 27.01.2021

[10] Alvares de Souza Soares, P., Geldmaschine Google – Wie lange noch? In: Handelsblatt, 25./26./27. April 2025, S. 26-27

[11] Knees, L., Warum Nutzer mehr für langsame KI zahlen. In: Handelsblatt, 31. März 2025, S. 24-25

[12] Holtermann, F., Schimroszik, N., Die Roboter kommen! In: Handelsblatt, 3./4./5. Januar 2025, S. 44-48

[13] O’Reilley, C., Tushman, M., Lead and Disrupt – How to Solve the Innovator‘s Dilemma, Stanford Business Books 2016

[14] Bomke, L., et al., Europa will eigene KI-Factories bauen. In: Handelsblatt, 9. April 2025, S. 6-7

[15] Servatius, H.G., Gestaltung von innovativen Stakeholder-Ökosystemen. In: Competivation Blog, 10.01.2023

[16] Brandenburger, A.M., Nalebuff, B.J., Co-Opetition – A Revolutionary Mindset That Combines Competition and Co-Operation, Bantam 1996

[17] Holzki, L., AMD schließt Partnerschaft mit der Industrie. In: Handelsblatt, 30. Januar 2025, S. 24

[18] Servatius, H.G., Die Ressourcen-Plattform mit agilen Teams als neue Organisationsform. In: Competivation Blog, 12.01.2021

[19] Doerr, J., Measure What Matters – How Google, Bono and the Gates Foundation Rock the World with OKRs, Portfolio/Penguin 2018

[20] Kaplan, R.S., Norton, D.P., Balanced Scorecard – Translating Strategy into Action, Harvard Business School Press 1996

[21] Bomke, L., Höppner, A., Nur wenige Unternehmen messen ihre KI-Initiativen. In: Handelsblatt, 16. Januar 2025, S. 21

[22] Gouldsberry, M., The Pivotal Role of AI in Performance Management, 11. Januar 2025

[23] Kaufmann, Servatius, a.a.O, S. 203ff.

[24] Servatius, H.G., Auf dem Weg zu einem neuen wirtschaftspolitischen Narrativ. In: Competivation Blog, 16.05.2022

[25] Servatius, H.G., Prozessorientierte KI zur Produktivitätssteigerung. In: Competivation Blog, 12.03.2025

[26] Servatius, H.G., KI und die Zukunft der Management Education. In: Competivation Blog, 09.04.2025

[27] Servatius, H.G., Nachhaltigkeitsorientiertes strategisches Management. In: Competivation Blog, 15.08.2024

[28] Servatius, H.G., Resilienzorientiertes strategisches Management. In: Competivation Blog, 15.03.2024

AI and the future of management education

09.04.25

The education and training of managers has gone through various phases in the past. At present, the negative consequences of too narrow a focus on theory are becoming increasingly clear and the idea of management as a dynamic profession that combines theory and practice is emerging. At the same time, artificial intelligence (AI) is changing learning and management education. Key drivers of this evolutionary change are disruptive learning ecosystems that are challenging the traditional players. The journey into this near future begins with an outline of the current patterns of success for a career in management.

In this new article in our series of blog posts on AI, I look at the present and future development of management education, in which artificial intelligence plays an important role.

Success patterns for a management career

In past decades, the proven success pattern for a turbo career in management consisted of the following steps:

Bachelor’s degree in engineering or information technology
short orientation phase in a reputable company
MBA at a renowned business school
further years of training in consulting, investment banking or private equity
change to a staff unit for corporate development and
rapid promotions on the way to management level.

However, this pattern of success is associated with two obvious disadvantages:

The limited work-life balance in professional services companies, which are considered to be labor-intensive, and
the cost barrier of MBA programs at well-known business schools.

For example, the cost of a one-year MBA course at one of Europe’s elite universities is between 80,000 and 100,000 euros. Added to this is the high cost of living at locations in metropolitan regions. In the USA, where most MBA programs last two years, the costs, including accommodation and health insurance, can add up to 250,000 euros.¹Applicants must therefore carefully consider whether such an investment and the associated loss of income is worthwhile.

In addition, starting in the USA and with a considerable delay, a second pattern of success has also emerged in Germany, albeit one that is associated with even greater risks. This path leads from an idea to the founding and scaling of a start-up. The players are usually creative, interdisciplinary teams that work on new business models with great commitment. A characteristic of successful founders is that they are also deeply involved in details and are familiar with important processes. There has long been no shortage of such talent in Germany, but there are deficits in the innovation policy framework. The situation has improved in recent years. But there is still a lot to be done.

One reason for the success patterns outlined above is that management science has become increasingly open to the technical sciences. Training in traditional engineering and the newer information technology qualifies students for a specialist career. However, a management career also requires skills in strategy, innovation, marketing, production, finance, organization and personnel management. An interdisciplinary management course teaches all of this. In addition, management education trains the skills that link to the technical sciences. An important field of application is the design of new business models.²

The pioneer of this development in Germany is the University of Stuttgart, where I have been teaching as an honorary professor since 1994 following my external habilitation. In Stuttgart, the technical-oriented business administration course was introduced as early as 1974, thus focusing on the connectivity of business and technology. The following diagram illustrates the segmentation of disciplines.

Less obvious than these patterns of success is a shift in focus in the content of management education and training. This development affects both business schools and traditional technology and business-oriented courses. One important point of criticism is their excessive focus on theory.

Too strong a focus on theory and its consequences

In an article published in the Harvard Business Review in 2005, the renowned US professors Warren Bennis and James O’Toole criticize a misguided development at American business schools. Triggered by funding from foundations, a stronger emphasis on scientific rigor in the form empirical treatment of relatively limited issues began there at the end of the 1950s. Publication in scientific journals became decisive for a career as a professor. Practical experience and the relevance of research fell by the wayside, playing an increasingly minor role in appointments^(3.

Unfortunately, the same development has been taking place at German universities for a long time. The result is a series of negative consequences:

Teaching has become less important than research
the professors bring less practical experience to their teaching and research
there are hardly any up-to-date textbooks, especially in dynamic subjects such as innovation management
research is dealing less and less with complex topics of practical relevance that are difficult to access using empirical methods
the target audience for the results of this research is primarily other researchers, while practitioners rarely read articles from scientific journals
many graduates are not immediately employable as project managers in demanding practical projects, even after completing a doctorate.

The major beneficiaries of this development have long been management consultancies and the consultants working there, who fill the gap left by business schools and universities. These winners distinguish themselves with their ability to tackle new, complex problems in a practice-oriented manner. As a result, consulting has become an important next step in professional development and a career accelerator. For clients, this causal chain has unfortunately increased their dependence on consultants.

It is astonishing that this development has not yet been analyzed and evaluated more critically by corporate practice. We therefore want to explore the question of what an approach to improvement could look like. To do this, management would have to transform itself into a dynamic profession.

Management as a dynamic profession

The education and training of managers is characterized by the three dimensions of theory, practice and the dynamics of change. Particular challenges lie in the combination of theory and practice as well as the increasing dynamics of the corporate environment. The idea of management as a dynamic profession is based on the development of management education and training.

This development has taken place in various phases. The emergence of management as an interdisciplinary field has a long history. The success of many German hidden champions in recent decades and also the phenomenal rise of digital companies from the USA is primarily based on managers with the ability to combine technical and business expertise.

The early degree programs made an important contribution to this connectivity in the first phase. With their pronounced focus on practice, prospective managers were very well prepared for the challenges of working in organizations. This phase could be placed under the motto of the great social psychologist Kurt Lewin: „Nothing is as practical as a good theory“. This implies that an excellent professor can be expected to have extensive practical experience in addition to academic achievements.

In a second phase of management education, the academization already outlined took place in the sense of restricting research to empirical approaches for dealing with limited issues. The majority of dissertations and habilitations now follow this path, which rarely leads to far-reaching management innovations.

Today, new impetus usually comes from professors who are also active in executive education and consulting or even as company founders. The results are often published in practice-oriented journals such as the Harvard Business Review and books by international publishers. Although this type of professor is in the minority in terms of numbers, they make a significant contribution to innovations in management theory.

A reorientation of management education is emerging in a third phase. This phase takes account of the highly dynamic nature of change and creates a stronger link between theory and practice. The result is a management that sees itself as a dynamic profession. The aim is to create innovative content for research and teaching in order to meet the many new challenges.

The following are important criteria for management activities to become a profession:

Qualified training with a degree
practical experience with demanding tasks
ethical standards and possibly also
a form of regulation, e.g. of AI.

In this respect, management’s claim to be a profession is currently only partially fulfilled. However, management can learn from other application-oriented sciences such as medicine.

Learning from medical training

No sensible person would allow themselves to be treated by a doctor with little practical experience who has never worked in a hospital. In order to ensure that medical training is not just theoretical, chief physicians at renowned clinics often work as professors at universities alongside their practical work. Part of the training takes place in teaching hospitals and overall the system of universities and clinics is much more permeable than the training and further education of managers.

The more practice-oriented universities and the dual study programs offered by private business schools do offer an alternative to studying at universities. However, most providers only have a limited right to award doctorates. This means that the contribution made by doctoral students to research is largely missing.

In view of the universities‘ strong desire to maintain their position, the best that can be expected from them in management education and training is change in small steps. A first step could be to integrate theoretically qualified managers and their companies much more closely into teaching and research. A long-established approach to this is action research.⁴This involves interdisciplinary teams designing innovative practical projects, publishing the results and making them accessible for teaching purposes. This has long been a common approach in management consulting. Both theoretically oriented professors and practitioners would benefit from such cooperation between universities and companies.

One exciting question now is what new impetus for management education will come from artificial intelligence (AI).

AI is changing management education

According to a study by the British „Times Higher“ magazine, thirteen of the world’s best AI universities are in the USA, three in the UK, two in China, one in Switzerland and one in Singapore. So there is still „room for improvement“ for the countries of the European Union. Chinese companies see 2025 as a key year for AI applications. This is why China is increasingly focusing on specific industry models in addition to large language models.⁵This is a strategy that Germany could learn from.

In addition to industry-specific models, AI is also becoming increasingly important for training and development. Almost all companies are faced with the challenge of upskilling and reskilling their employees successfully and cost-effectively.

This also applies to management training and development. AI has a profound impact on four aspects of management. It changes:

Research and teaching as well as learning technologies
all functions and business processes of organizations
all economic sectors and the public sector as well as
management as a developing profession.

The practical skills of managers and employees who use AI as a tool are of great importance here.

Engineering and management courses are now facing the challenge of integrating the fundamentals and applications of artificial intelligence into their curriculum. The learning content is developing very dynamically and individual professors therefore often feel overwhelmed by this integration task. According to the European Union’s AI Act, companies are obliged to train employees who use AI systems in how to use them. Many companies are not yet aware of this. In July 2024, the German association Bitcom asked German managers about their attitude towards AI. According to the survey⁶
– 29% of respondents are cautious and 16% skeptical, but still
– 46% are keen to experiment and 9% are even enthusiastic.

Artificial intelligence is therefore both a learning content and a learning tool that is changing university teaching and practice.

In this context, it is remarkable how dynamically learning technologies have developed over the past two decades. We therefore want to take a look at the possible future of corporate learning.

Development towards AI-based corporate learning

Since the turn of the millennium, the development of learning technologies used by organizations has gone through the following four phases:

E-learning with Learning Management Systems (LMS) as a platform
further development of the LMS into a more competence-oriented learning in the sense of a personnel development system
digital microlearning with Learning Experience Platforms (LXP) and the increasing importance of videos, as well as
AI-based learning platforms that generate individualized content relatively independently.

In each of these phases, LearnTech providers have disappeared from the market and new ones have emerged. Providers that rely on AI include Absorb, Arist, Docebo, Growthspace, LearnUpon, Sana Labs, Uplimit and Work Ramp⁷

Planning the use of learning technologies in organizations is often not done top-down, but bottom-up by HR management and information technology. Cooperation with start-ups plays an important role here

The young Berlin-based company Doinstruct uses AI to train „desk-less“ employees. Gartner estimates this target group at 2.7 billion people worldwide. It is twice as large as the number of people who work at a desk. The company uses AI to generate short educational videos for „frontline workers“, who often have neither a laptop nor an email address. So far, Doinstruct has produced more than 250 training videos. Companies can personalize this offer. The basis for the generated videos is provided by well-known AI providers. Doinstruct’s experts select the content and scriptwriters prepare it. The company has developed software that sends the log-in via SMS. WhatsApp chats are also being considered. The content is translated into 25 languages.⁸

Procedure for AI-supported learning in management

The Doinstruct example shows what AI-supported learning in management can look like. It consists of seven steps that are adapted to the respective situation.

The first step is an AI-supported summary of existing learning content. This step requires prompting supported by reasoning models and a critical analysis of the AI-generated content. It is not intended to replace teachers, but above all to increase their productivity.

In step two, project-oriented action learning by universities and companies creates innovative, practice-oriented learning content. In this way, new knowledge and skills are incorporated into teaching.

The third step is about combining the basics with new content and structuring it into microlearning. These short sequences enable the modularization of learning programs.

A multimodal, didactic preparation is then the task of step four. In this way, a professional combination of text, sound and image is achieved, which is combined with practical exercises. These first four steps can be used for specific management disciplines and industries.

In the fifth step, there is the option of AI-supported individualization of the learning content. This allows it to be adapted to the specific situation of a company and individuals.

In the sixth step, innovative learning technologies are used to further develop learning platforms that enable scaling. Most companies will master this step together with LearnTech partners.

This lays the foundation for step seven, which focuses on AI-supported continuous improvement. Disruptive learning ecosystems play an important role in the implementation of such an approach.

Change through disruptive learning ecosystems

A disruptive learning ecosystem is a network of partners from politics, science, business and the learning community whose disruptive effect comes from innovative approaches and barriers that are difficult to overcome by traditional players.⁹One advantage of the emerging AI-based stakeholder ecosystems is better cooperation between the partners.

Such a learning ecosystem consists of the following players:

Politics that create a framework that promotes innovation
universities, consultants, trainers and authors who conduct research and offer learning content
traditional and innovative education providers that market learning content
AI and LearnTech providers who design learning platforms and
organizations and individuals in the dual role of customer and provider of learning content and data.

The opportunity for a more resilient Europe lies in taking a pioneering role in such new educational systems with trustworthy AI solutions. The goal is connective management education for the changing world of work in the AI age.

Connective management education for the AI age

An important characteristic of the fifth development stage of connective strategic management ¹⁰is personnel development that takes into account the specific requirements of the changing world of work in the AI age. This connective management education combines elements in the following fields:

Theory and practice (A)
training and further education (B)
specialization and interdisciplinarity (C) and
human strengths and artificial intelligence (D).

The harmonization of these elements is an important design approach within the framework of Strategy 5.0, which enables new competitive advantages. The basis for this is a changed attitude that is characterized by a future-oriented spirit. Such an attitude combines optimism with openness and curiosity with the joy of experimentation to create a mental agility that helps to recognize and exploit opportunities.¹¹

The first of these four fields is a better combination of theoretical principles and their practical application. This involves both training for a changed world of work and the further training of many people in a world of work that is changing as a result of AI. AI is giving rise to new professional specializations. At the same time, the importance of a more interdisciplinary approach to management education is increasing. Overall, it is important to combine human strengths in teaching and research, such as a forward-looking approach, with the use of AI as a tool.

This fourfold connectivity has now become one of our focal points in research, teaching and consulting.

Conclusion

Universities should strive for a stronger link between theory and practice in the education and training of managers. Medical training provides inspiration on the path to management as a dynamic profession
Artificial intelligence (AI) is not only changing the world of work, but also management education. AI is both a learning content and a tool
LearnTech providers and user companies are working on AI-based learning platforms, which are the focus of a new development phase in corporate learning
Disruptive learning ecosystems challenge traditional players, but also represent an opportunity to realign the European education system
Connective management education for the AI age combines elements in various fields.

Literature

[1] von Elm, K., Beware of lofty expectations. In: Handelsblatt, March 07/08/09, 2025, p. 36-37

[2] Servatius, H.G., Evolution of strategic management. In: Competivation Blog, 28.06.2024

[3] Bennis, W.G., O´ Toole, J., How Business Schools Lost Their Way. In: Harvard Business Review, May 2005, pp. 96-104

[4] Servatius, H.G., Generative AI and Mass Customized Action Learning. In: Competivation Blog, 28.08.2023

[5] Bomke, L., Gusbeth, S., Who will win the AI race? In: Handelsblatt, March 10, 2025, p.18

[6] Burkhardt, K., Anyone working with AI systems must be trained beforehand. In: Handelsblatt, March 18, 2025, p. 26-27

[7] Bersin, J., The $ 340 Billion Corporate Learning Industry is Poised for Disruption, March 23, 2024

[8] Schimroszik, N., Doinstruct trains employees with AI. In. Handelsblatt, March 19, 2025, p. 27

[9] Servatius, H.G., Designing innovative stakeholder ecosystems. In: Competivation Blog, 10.01.2023

[10] Servatius, H.G., Strategy 5.0 for mastering the new challenges. In: Competivation Blog, 28.06.2022

[11] Pferdt, F.G., Radikal besser – Entfache den Zukunftsgeist, der in dir steckt, Hamburg 2025

« Older Entries

Designing trustworthy high-performance systems

AI as a tool for strategic management

KI als Werkzeug für das strategische Management

AI and the future of management education

Neueste Beiträge

CONNECTIVE MANAGEMENT