Anna Drobakha, Global Digital Business & AI Transformation Director at Groupe SEB.
getty
There's a pattern I keep seeing across organizations.
In conversations with leadership teams, the ambition around AI is clear. There's energy, investment and genuine belief that it will reshape how the business operates. Yet when we look at actual impact, progress often feels slower than expected.
AI pilots are everywhere, prototypes are promising and leadership conversations are active. Still, real business outcomes remain uneven.
The common explanation is technical: not enough data, not the right models, not the right tools. In practice, that’s rarely the real issue. The deeper constraint is organizational.
Over the past two years, the accessibility of AI has changed dramatically. Capabilities that once required specialized teams are now available across functions. Experimentation is easier, and entry barriers are lower.
Yet outcomes remain uneven. In many cases, the limiting factor is whether the organization is able to work in a way that allows AI to be used effectively.
Most companies are still trying to fit AI into structures that were designed for a different era—predictable processes, centralized decision making, clear functional boundaries and controlled execution. These models work well for efficiency and risk management. They struggle when speed, iteration and adaptation become critical.
AI introduces a fundamentally different dynamic into organizations.
To create value, teams need to move quickly to test ideas, learn continuously, iterate based on real feedback and combine perspectives across functions. Progress depends far less on perfect planning and far more on fast learning cycles.
This is exactly where friction appears. In many organizations, even simple experiments require multiple approvals, alignment across layers and coordination between siloed teams. By the time something's validated, the momentum is already lost.
What looks like "AI not delivering impact" is often a reflection of how decisions are made, how ownership is structured and how quickly teams are allowed to move.
Across different environments, the patterns are surprisingly consistent. AI initiatives often start with energy but remain confined to innovation or pilot teams. They struggle to move into day-to-day operations and are dependent on a small group of experts instead of broader capability.
At the leadership level, there's alignment on the importance of AI but less clarity on how to operationalize it. The result is a gap between ambition and execution—not because organizations lack intent but because the system they operate in slows them down.
The organizations starting to see real impact are the ones making subtle but important shifts in how they operate. They reduce unnecessary decision layers, push ownership closer to the teams doing the work and enable structured experimentation instead of waiting for perfect plans.
Importantly, this is about redefining control. Control moves from approving every step to setting clear direction, defining guardrails and ensuring accountability. Within those boundaries, teams are able to move faster. This creates a different kind of organization: one that's still aligned but far more adaptive.
What's emerging is a gradual shift in how organizations are designed.
In different forms, across industries, I increasingly see similar patterns—particularly in more advanced organizations. Teams are given more ownership over decisions. Experimentation becomes part of how work gets done, not an exception. Leaders spend less time approving and more time setting direction and context.
A few years ago, these ideas were often discussed at the edges, associated with progressive or experimental organizations. What's changing now is where they're showing up.
Increasingly, even large and traditional companies are starting to adopt elements of this model as a response to a simple constraint: Their existing structures are too slow for the environment they operate in. What used to be considered progressive organizational design is quickly becoming a practical requirement.
Many leadership discussions still focus on how the organization implements AI. A more useful question is: “What in our organization makes it hard for AI to work?”
This shifts the conversation from tools to systems and surfaces practical constraints: Where does decision making slow down? Where do experiments get blocked? Where is ownership unclear? Answering these questions often reveals that the real work of AI transformation is organizational.
AI is often described as a technology shift. In reality, it's exposing something deeper. It's making visible how organizations actually function, how quickly they learn, how decisions are made and how responsibility is distributed.
The companies that will scale AI successfully are evolving how they operate by becoming more intentional about where control is needed and where it holds them back.
In an environment defined by speed and uncertainty, the advantage comes from having an organization that can learn and act faster than the rest—and that's a leadership decision.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。