Bottom line
The modern enterprise org chart was designed for an economy where information was scarce and expensive. Hierarchies existed to filter, aggregate, and route information upward to the people with the authority to act on it. Middle management was, in large part, an information processing layer.
What changes when information becomes cheap.
AI doesn't just automate tasks — it makes analysis, synthesis, and structured reasoning dramatically cheaper. The information processing function that justified many organizational layers can now be performed faster and more consistently by AI systems. This doesn't eliminate the need for human judgment; it relocates where judgment is most valuable.
The decision rights question.
AI-first org design starts from decision rights, not from reporting lines. Which decisions require human judgment, and at what level? Which decisions can be AI-assisted with human approval? Which can be fully delegated to structured AI systems? The answers to these questions generate very different organizational structures than the ones inherited from pre-AI knowledge work.
More on Enterprise Transformation — every week.
Join 2,400+ leaders getting the decision intelligence briefing.
Flatter but more accountable.
The AI-first organizations we see emerging share a structural characteristic: they're flatter, but with much more explicit accountability at every level. Because AI can handle more of the analytical work, fewer layers of review are needed. But the humans who remain in the decision chain carry clearer, more defined responsibility — and they're supported by systems that create better records of what was decided, why, and by whom.
The new role of middle management.
The managers who thrive in AI-first organizations aren't the ones who were best at aggregating and filtering information. They're the ones who are best at framing decisions clearly, designing workflows, developing people, and exercising judgment where structure is insufficient. This is a genuinely different skill set, and organizations that don't invest in developing it will find their middle layers becoming a drag rather than an asset.
Redesigning for AI isn't primarily a technology project. It's an organizational design project that happens to use technology as its enabler.
Research context
BCG found that only 26% of corporate transformations created value in both the short and long term over the two decades studied.
Boston Consulting Group: Five Truths (and a Lie) About Corporate Transformation (2024)This independently published statistic provides context for the topic; it does not substantiate every claim in this essay.
More on Enterprise Transformation, weekly.
One essay. Every week. No noise.