Who Leads Artificial Intelligence in the Organization?

Who should really lead the AI revolution inside the organization? Behind this question lies a much deeper transformation - one that challenges organizational structures, managerial responsibility, and the way organizations think about work, knowledge, and decision-making.

Prefer to Listen? Tune in to the Podcast

From Technological Leadership to an Organizational AI Office

Over the past two years, almost every organization has been asking itself similar questions: Who should lead the field of artificial intelligence?

Where is the right place to position it within the organization? Is this the responsibility of information systems, data, innovation, business management, or perhaps an entirely new body?

It is interesting to see that there is no single unified model today. In practice, a wide variety of organizational structures can be found worldwide: in some organizations, AI leadership sits under the CIO; in others, under data and analytics units; sometimes under innovation or digital transformation; and even within the business units themselves.

Nevertheless, when looking at the broader market - and especially the enterprise market - a clear pattern emerges: in most cases, at least in the first stage, AI leadership grows out of the computing, digital, or data worlds.

The reason for this is simple and logical. The early stages of organizational AI were primarily concerned with infrastructure, models, information security, integration with core systems (Data Governance), and technological capabilities. Therefore, naturally, IT professionals were the first to raise the flag.

But as usage expanded, it became clear that the real challenge is not only technological. Organizations discovered that successful pilots do not necessarily produce business change; that an excellent system, if not integrated into work processes, is barely used; and that employees do not change their behavior patterns simply because a new tool has been added.

And this is where the truly interesting organizational evolution begins.

If we summarize the trends emerging in the world today, a fairly consistent development model can be identified between the level of organizational maturity and the place where the AI domain sits:

Leading department	Maturity Stage
IT / Data	Early Stage
Digital / Innovation / Transformation	Scaling
AI Office / Federated Model	Enterprise-Wide
Business ownership alongside central Governance	Mature Organizations

In the first stage, AI is managed primarily by IT, Digital, or Data & Analytics. The goal at this stage is to build capability: to select platforms, manage risks, connect infrastructure, and begin experimenting. This is an important and necessary stage; without a stable technological foundation, it is very difficult to advance.

But after the initial enthusiasm comes the understanding that the question is no longer “Do we have intelligence?” but rather “Are we truly working differently because of it?”

And then the focus shifts away from the technology itself and toward adoption, policy, measurement, process change, AI literacy, managerial accountability, and role re-engineering.

The discussion gradually moves from technological capabilities to organizational capabilities.

This is where we see a very accelerated trend over the past two years: the establishment of an AI Office or AI Transformation Office. It is not always a large unit; sometimes it is a relatively small body, but one with significant cross-organizational responsibility. Its role is not to replace information systems, but to integrate the organizational layer of the AI revolution.

In practice, the AI Office typically operates across several key domains to ensure optimal, responsible intelligence. It does so through a computational infrastructure (supported by information systems), a process infrastructure (policies, standards, leadership training, and the reinforcement of overall literacy), and a distributed organizational infrastructure of local AI leaders within the business units, who create local business leadership.

This is perhaps the most important development. Advanced organizations understand that AI cannot be managed solely from headquarters. A layer of champions is therefore being built - business leaders and local managers who take responsibility for the applications, for their adaptation to work processes, and for the re-engineering of the roles themselves. Ownership gradually transfers from technology to business operations.

At this point, it is also important to clarify the role of the AI Office. Such a body is not meant to become a “project unit” that manages or approves every initiative in the organization. It is generally more appropriate for it to focus on only two types of projects:

On one hand, an innovation lab that examines new technologies, runs experiments, and identifies future opportunities;

On the other hand, cross-organizational horizontal projects require a unified standard, shared infrastructure, or coordination between many units. The remaining initiatives should gradually pass to the ownership of the business units themselves.

It is also important to add one caveat: this model is less suited to “pure” technology organizations. In SaaS or AI-native companies, artificial intelligence is neither a separate transformation layer nor merely a tool for improving organizational work - it is part of the core product and R&D. In such cases, leadership will often fall under Engineering, Product, or Research.

The Knowledge Management Angle

Here, the picture is complex and interesting. On one hand, the AI Office is not “the new knowledge management,” and in most organizations, it will not sit under knowledge management. On the other hand, it is hard to ignore that many of the challenges that AI raises are, in essence, knowledge challenges: how to make expertise accessible, how to enable the reuse of insights, how to integrate human knowledge with AI, and how to turn knowledge into better business action.

Therefore, in more mature organizations, knowledge management professionals are beginning to find their place in domains such as AI literacy, the integration of human knowledge in AI processes, communities of practice, Governance of content and knowledge, and support for behavioral change, not as the sole owners of the field, but as a significant part of the new organizational capability being built around it.

It is possible that in the near future, we will not speak at all about “AI adoption” as a separate field. Just as there is no “Internet office” or “cloud computing department” today, so too will AI gradually become a natural layer within organizational work itself.

And perhaps the real question will no longer be “Who leads AI in the organization?” but rather “Does the organization know how to redesign itself around intelligence - human and artificial alike?”