Over the past few months, I’ve seen the same situation play out time and again at various companies: A board of directors adopts an ambitious AI strategy. Millions are invested, project teams are formed, pilot projects are launched, and employees are trained. At first glance, everything seems to be moving forward.
Six or twelve months later, however, the results are sobering. Technically, the systems are working; productivity is increasing in some areas, and routine tasks can be completed more quickly. However, the actual transformation has yet to materialize. Decisions still take weeks, innovation occurs only sporadically, and promising ideas get lost in coordination loops and approval processes.
Many executives even view artificial intelligence as an additional burden. They are expected to adopt new technologies, while existing structures, reporting lines, and decision-making processes remain virtually unchanged.
So when I look for the causes, I rarely focus on the software. Much more often, my attention turns to the executive suite.
At first glance, this sounds provocative. After all, board members facilitate the necessary investments, set strategic goals, and drive change. Nevertheless, that is often where the real obstacle lies. Many board members unconsciously defend a leadership model that has made them successful. Almost all of today’s top managers rose through the ranks at a time when leadership primarily meant consolidating information, controlling decisions, and concentrating responsibility in the hands of a few individuals. Leadership strength was closely linked to a knowledge advantage, decision-making authority, and personal control.
This model made sense for a long time. It emerged in a world where information was expensive, slow, and incomplete. Anyone who had access to the relevant data and could synthesize it into a clear picture had a decisive advantage.
Artificial intelligence is radically changing this foundation.
Today, information is available almost everywhere at the same time. Analyses are generated in real time, scenarios can be simulated within minutes, and employees have access to knowledge that was once reserved for senior management.
The key bottleneck, therefore, is no longer access to information. It lies in the willingness to redistribute responsibility.
This is exactly where the dilemma begins. Many board members want quick decisions, innovative teams, and leaders with an entrepreneurial mindset. At the same time, they cling to processes that force every major decision back up the chain of command. Strategic approvals, budget decisions, prioritizations, and even operational issues go through numerous rounds of coordination.
These processes are intended to ensure safety. In practice, however, they often lead to delays, uncertainty, and a gradual erosion of accountability.
Those who prepare decisions but are rarely allowed to make them on their own learn, over time, to pass on responsibility. Those who must expect to be asked for clarification, provide justifications, or make retroactive corrections whenever there is a deviation develop risk-mitigation strategies instead of an entrepreneurial mindset. Therein lies one of the greatest contradictions of our time:
Companies invest millions in technologies designed to speed up decision-making, yet they do not trust their own executives enough to actually let them make those decisions where the knowledge lies.
AI systems process complex information within seconds. This is followed by an approval process lasting several weeks. The technology delivers scenarios in real time, while the organization waits for the next regular meeting, the next steering committee meeting, or the next board meeting. Artificial intelligence accelerates the analysis. The organization slows down the decision-making process. Technological excellence gives rise to organizational inertia. This inertia often remains hidden for a long time. The systems are running, projects are presented as successes, and initial efficiency gains can be demonstrated. At the same time, the underlying structures remain untouched. Decision-making authority remains unclear, responsibility is delegated only to a limited extent, and executives are guided more by internal expectations than by market demands. The result is a form of digital modernization in which old management paradigms are simply equipped with new technology.
That is why a key insight keeps being confirmed time and again in my projects: The most successful companies do not necessarily have the most advanced AI systems. What matters most is their willingness to rethink responsibility.
Technology can provide information, reveal connections, and improve decision-making options. However, its potential remains limited as long as every relevant decision must be approved centrally. AI can only realize its full potential when it is accompanied by a change in the leadership structure.
This inevitably changes the role of the Executive Board.
In the future, executives will be able to measure their success less and less by how many important decisions they have made personally. Their real task is to create a system in which as many good decisions as possible can be made independently of their own presence.
The Executive Board is evolving from the top decision-maker to the architect of a decision-making space.
It establishes the framework, defines guidelines, and ensures that responsibility can be assumed where expertise, experience, and familiarity with the relevant topic exist. Such a decision-making environment requires clarity. Teams need to know what goals apply, what risks are acceptable, within what limits they are permitted to act, and when a decision must be escalated. Without guidance, effective autonomy cannot develop; instead, uncertainty prevails.
The shift in responsibility therefore does not mean that the Executive Board is stepping back. It requires a more precise form of leadership. Goals must be formulated more clearly, priorities set more consistently, and conflicts resolved more quickly. The Executive Board continues to bear responsibility, though it is now more accountable for the quality of the entire system than for any single decision.
Oversight is also changing in nature. The board is focusing less on reviewing individual decisions and more on examining the conditions under which they are made. Is information accessible? Are roles and responsibilities clearly defined? Do incentives encourage independent action? Can the organization learn from its mistakes?
The crucial question is then no longer just:
“Who approved this decision?” The question should be: “Why was our system able to produce a good decision—or why did it fail to do so?”
For me, that is precisely where the true significance of AI-Augmented Leadership lies. This leadership model describes far more than just the use of artificial intelligence. It describes a new approach to responsibility in a world where knowledge has become ubiquitous. The better AI provides information, the less leadership can define itself through control over information. Guidance, trust, and judgment become more important. Modern leadership must ensure that people can contextualize information, weigh the consequences, and become capable of taking action within a clear framework.
AI does not replace either experience or human judgment. It reveals patterns, highlights connections, and opens up new perspectives. The responsibility for making decisions remains with humans. That is precisely why the quality of leadership continues to grow in importance.
AI-Augmented Leadership requires a new form of trust. This trust is based on clear roles, transparent performance, and transparent decision-making principles.
Employees will only take on responsibility if they see that acting independently is actually encouraged. Decisions must be upheld as long as they were made within the agreed-upon guidelines. Anyone who delegates responsibility and then takes it back at the first sign of an uncomfortable consequence will quickly destroy any sense of initiative.
The organization then learns that autonomy is only valid as long as the results meet the expectations of top management. This leads to conformity, caution, and internal politics. Innovation requires people who can handle uncertainty, take calculated risks, and stand by their decisions.
This requires a quality that is rarely mentioned in many discussions about digitalization: humility.
The humility to recognize that good ideas do not necessarily originate at the top of an organization. The humility to accept that, with the support of intelligent systems, employees are often better positioned than top management to make the right decision in many situations. And the humility to understand that modern leadership does not have to know the best answer to every question itself. Its task is to create conditions under which the best answers can emerge within the company.
For many board members, this is a personal challenge. Those who have been rewarded for decades for making quicker judgments, knowing more, and having the final say at the decisive moment may easily perceive the relinquishment of decision-making power as a loss of significance. In fact, the board’s role is shifting from making individual decisions to shaping the organization as a whole.
An executive board that has to answer every important question on its own will inevitably become a bottleneck. An executive board that has created an effective decision-making system multiplies its impact. Artificial intelligence will therefore not render executive boards obsolete. It will force them to redefine their role. The crucial question is not merely whether companies successfully deploy AI. What matters is whether their leadership is willing to rethink power. In the future, power will be demonstrated less and less by centralizing decisions. It will be demonstrated by the ability to distribute responsibility in such a way that the organization can act faster, smarter, and more independently.
That is precisely where it is determined whether technological innovation will translate into organizational innovation. And perhaps that is the biggest difference between the winners and losers of the AI era: the quality of their leadership.
by Frank Rechsteiner