The 'Great Flattening'—AI-driven organizational flattening—is a structural transformation actively reshaping modern enterprises. While promising unparalleled efficiency and agility, it simultaneously creates significant risks to institutional knowledge and talent development pipelines.
A decade of organizational bloat, with many companies adding one to three management layers and creating complex, slow, expensive hierarchies, is now being reversed by AI. Firms like McKinsey & Company confirm this trend, noting corporate leaders are using AI to redesign org charts into "agentic organizations." Understanding this paradigm's potential and operational hazards is crucial.
AI's Role in Reshaping Team Structures and Hierarchies
The 'Great Flattening' is a tangible restructuring of corporate teams, removing management layers, reassigning or eliminating junior and support roles, and embedding AI for tasks previously handled by humans. AI's capacity to automate functions that necessitated managerial oversight is the catalyst.
For decades, the span of control for any given manager was limited by human cognitive capacity. A leader could only effectively oversee, coordinate, and synthesize information for a finite number of direct reports. AI shatters this limitation. According to reporting from Startup Fortune, AI gives leaders "more of a superhuman capacity to manage across bigger scopes," enabling a radical expansion of their oversight. This newfound capacity is built on several key AI-driven functions:
- Automated Coordination: Middle managers reportedly spend approximately 40% of their time on coordination tasks—scheduling, tracking progress, and facilitating communication. AI agents can now perform many of these functions seamlessly, freeing up human capital for higher-value work.
- Enhanced Data Synthesis: Instead of relying on a chain of managers to distill frontline information for executive review, AI can aggregate real-time data, identify trends, and generate reports instantaneously. This allows senior leaders to have a direct, unfiltered view of operations.
- Task Execution: The model is already in practice. McKinsey itself is reportedly deploying thousands of its own AI agents to support consultants with research, data analysis, and even building presentation decks. As Rob Levin, a senior partner at McKinsey, explained to Fortune, the old model is giving way to a "more flat network of human teams supervising AI agents."
Moderna reportedly merged its technology and human resources departments, partly driven by AI use for HR support and junior-level tasks. This consolidates functions, with technology picking up slack for a leaner human structure.
The Counterargument: A Redefinition, Not an Elimination
While some suggest a future with vast, automated teams overseen by senior executives, middle management is not obsolete but redefined. Its core function evolves from coordinating tasks to orchestrating human and artificial talent.
Tristan L. Botelho, an associate professor at the Yale School of Management, articulated this perspective to Fortune, stating, "I don’t think middle management is going to be erased. I think it’s going to just redefine how managers think about their role within the organization." In this vision, the manager's value shifts entirely. With AI handling the administrative and analytical burdens, the manager's primary responsibilities become those that are uniquely human: coaching, mentoring, fostering psychological safety, and steering team culture. They become the essential human interface in an increasingly automated system, ensuring that strategy is executed with context, empathy, and creative judgment.
While this is an optimistic and strategically sound vision, I believe it overlooks a critical flaw in its own logic. It presumes a steady supply of leaders equipped with these sophisticated orchestration skills. Yet, the very act of flattening removes the traditional proving grounds where those skills are honed. The junior and mid-level management roles that are being automated away were the sandboxes for developing future leaders. They were where aspiring executives first learned to navigate difficult conversations, manage team dynamics, and take responsibility for outcomes. By eliminating these rungs on the ladder, organizations risk severing their own talent pipelines, creating a future leadership vacuum that no amount of AI can fill.
Navigating the Challenges of AI-Driven Organizational Flattening
Leaders must recognize organizational efficiency does not equal long-term effectiveness. A flattened structure, optimized for speed and low overhead, risks becoming brittle and susceptible to challenges a traditional hierarchy mitigated. The most significant challenges are deeply human, not technological.
Consider the implications for institutional knowledge. AI models are exceptionally proficient at analyzing explicit data, but they cannot access what one report aptly called the "scar tissue" of an organization—the invaluable, hard-won lessons learned from past failures, near-misses, and market shocks. This tacit knowledge often resides within experienced mid-level managers who have seen strategies succeed and fail. Their departure leaves a void of context and wisdom that cannot be queried from a database. When the next unforeseen crisis hits, the absence of this human-held knowledge could prove catastrophic.
Furthermore, the erosion of mentorship pipelines presents an existential threat to sustainable growth. Culture and professional norms are not learned from employee handbooks; they are transmitted through observation, apprenticeship, and direct human interaction. In a flattened hierarchy where a handful of individuals oversee vast teams of people and AI agents, the organic, day-to-day opportunities for mentorship evaporate. Junior talent may execute tasks efficiently with AI assistance, but they will miss the crucial development that comes from watching a seasoned manager handle a difficult client, negotiate a complex deal, or lead a team through uncertainty. This creates a generation of tactically proficient but strategically underdeveloped employees.
What This Means Going Forward
The 'Great Flattening' is not a passing trend but the beginning of a fundamental reorganization of work. However, its current trajectory is unsustainable. I predict that while the initial push to flatten will accelerate over the next two years, it will be followed by a "Great Re-layering" as organizations confront the hidden costs of a depleted leadership pipeline and the loss of critical institutional memory. The pendulum will not swing back to the bloated hierarchies of the past but will settle on a more sophisticated hybrid model—one that leverages AI for efficiency while intentionally creating structures to foster human development and preserve knowledge.
To navigate this transition, leaders must first audit organizational knowledge: map vital, unwritten operational wisdom and design systems to capture and transfer it before restructuring. Second, re-architect mentorship for a flattened world. Replace disappearing manager-mentee relationships with deliberate alternatives like project-based coaching, formal sponsorship programs, and peer-to-peer knowledge-sharing networks.
Leaders must lead AI, not be led by it. AI systems are powerful analytical tools, not autonomous decision-makers; human accountability for strategic, ethical, and cultural judgment is essential. The 'Great Flattening' removes organizational friction but is not a strategy itself. Companies succeeding will flatten thoughtfully, augmenting people with AI, not just replacing them.
