Transforming With AI Requires More Than Tools: The Rise Of The Enterprise AI Operating Model

Alejandro Oses, CEO and cofounder at Rootstack. I lead digital accelerations for companies across all industries.

​The conversation in the offices of company executives in the past two years has revolved around what artificial intelligence tool they should buy. But that question already falls short, because adoption is no longer the focus. Now, the challenge for companies is for this adoption to truly reflect financial results.​

One of the most recent McKinsey studies conducted in 2026, with 1,719 participants from 97 countries, found that 44% of organizations already scale AI throughout the company and that 80% report improvements in their individual productivity.​

But the proportion that attributes an impact on profits (EBIT) to AI remains at 37%, the same as in 2025. Our teams work better, but our companies do not necessarily earn more.​

The gap is organizational, not technological​.

Seeing this panorama, it is evident that this individual gain in productivity does not necessarily translate into better financial performance: It is necessary to redesign processes and make changes to the operating model.​

Only 6% of respondents, the so-called “high performers,” attribute at least 5% of their EBIT to AI and describe that value as significant. In the previous edition of the study, these companies were almost three times more likely to fundamentally redesign their workflows.​

It is not a question of blindly affirming that the redesign of processes causes these results, but it is coherent since I give you this example: Implementing an AI assistant in a legacy process produces half-hearted efficiency, while rebuilding the process around AI capabilities produces competitive advantage.​

The AI ​​operating model changes the conversation​.

A Deloitte study reported that almost 75% of the executives consulted recognize that their operating model will have to change during the next 12 to 18 months to sustain the advancement of AI.​