AI transformation fails not because of technology - but because of organization, AI skills, and leadership

• High ambitions, low implementation: 62 percent of companies anticipate profound AI-driven changes to their operating models, yet only 38 percent have initiated the transformation
• It is not technology holding AI back, but rather organization, AI skills, and leadership: 49 percent cite a lack of AI skills as the biggest hurdle, while 37 percent consider their structures and processes unsuitable
• Missing foundations hinder AI scaling: 59 percent believe their leadership is insufficiently prepared, and 42 percent have doubts about their governance structures

Munich, July 2026: Companies are investing heavily in artificial intelligence (AI), yet a major economic breakthrough often fails to materialize. According to the new Roland Berger study, "The AI-First Organization – Turning AI power into enterprise performance," the reason lies not in the technology itself, but in outdated organizational models. The study surveyed 472 executives worldwide. The findings reveal that while many companies recognize the magnitude of the AI transformation, they have so far failed to adequately adapt their structures, decision-making processes, and leadership models.

The figures highlight a clear implementation gap: 62 percent of respondents anticipate major or radical changes to their operating models, yet only 38 percent have actually begun the corresponding transformation. At the same time, 59 percent feel their leadership teams are not sufficiently prepared for the impending changes. The study thus illustrates a widening gap between strategic ambition and operational reality.

Particularly noteworthy is that respondents identify people, skills, and competencies—rather than technology—as the primary obstacles to generating value through AI. 49 percent view this as the biggest barrier. Organizational structure and processes follow at 37 percent, with technological requirements ranking third at 34 percent. At the same time, many organizations lack the foundations for scalable AI adoption: 42 percent of companies are unsure whether they have the right governance structures in place, and 59 percent report that their leadership teams are inadequately prepared.

The study also clearly highlights what distinguishes leading companies—those that have already made successful strides in their AI transformation—from the rest: 65 percent operate on a shared technology platform to interconnect data, processes, and systems across departments, compared to just 18 percent of less advanced companies. Furthermore, 58 percent are building AI capabilities broadly across all functions, whereas only 14 percent of laggards are doing so. The gap is also evident in decision-making and collaboration: 88 percent of AI pioneers make decisions systematically based on data, and 73 percent work in cross-functional, agile teams.

"Many companies are making rapid progress with specific AI applications, yet the organization often fails to keep pace with this speed," says Cyrus Asgarian, Senior Partner at Roland Berger. "However, this is precisely where it is determined whether technological advances translate into sustainable impact. To achieve this, governance, data foundations, processes, and leadership must be strategically developed, and the operating model consistently aligned with AI."

"Now is the time for companies to establish the structural prerequisites for widespread AI adoption," adds Niels Kammerer, Partner at Roland Berger. "Clear responsibilities and an integrated data and technology platform are crucial for enabling faster, data-driven decision-making. Additionally, the concept of leadership must be redefined for a world where performance is delivered through a hybrid mix of employees and AI agents."

About the study

The study is based on a global survey of 472 executives from large and medium-sized enterprises. The survey was conducted between December 2025 and April 2026 and covers the following sectors: automotive, industrial goods, financial services and insurance, pharmaceuticals and healthcare, consumer goods, retail and logistics, and telecommunications and media. Geographically, the participants are distributed across Germany, Austria, and Switzerland (35%), the rest of Europe (30%), North America (15%), the Middle East (10%), and Asia and Oceania (10%).