The Role Nobody Redesigned: Middle Managers and AI | AGORÀ Blog

McKinsey research published in 2024, including the analysis Middle managers hold the key to unlock generative AI (McKinsey, 2024)positions middle management as a critical variable in enterprise AI deployment. The research quantifies the automation exposure: a significant share of the tasks that define middle management operations, data collection, performance monitoring, report synthesis, upward communication, can now be handled by AI systems at lower cost and higher frequency.

The automation case is documented. The organizational design response to it is largely absent.

What the employee data reveals

In October 2024, Gartner surveyed nearly 3,500 employees on management and AI. The finding that stands out: 87% believe algorithms could deliver fairer performance feedback than their current managers (Gartner via HR Executive, 2025).

This is not primarily a statement about algorithmic superiority. It is a signal that the feedback function, traditionally a core middle management responsibility, is being measured by employees against a different standard. The implication for role design is direct: as AI handles more of the data collection and monitoring infrastructure, the expectations placed on human managers shift toward judgment, interpretation, and relational capability. The role's center of gravity moves.

Most organizations have not moved the role description with it.

The strategy gap at the top

A Gartner survey of 197 C-level executives and senior business leaders, conducted in December 2025, found that only 27% have a comprehensive AI strategy in place (Gartner, 2026). Among those, just 20% believe their workforce is genuinely AI-ready.

The pattern these figures describe is consistent: organizations are deploying AI tools at the operational level while the structural framework for what changes at the management level remains incomplete. Tools arrive. Role definitions do not update.

This creates a specific governance problem in the middle of the hierarchy. The data is there, real-time, AI-generated, often more complete than what managers previously had access to. The question of what a manager is expected to do with that data, what decisions belong to them, what authority they carry in an environment where AI handles the first layer of analysis, that question is frequently unanswered.

The skills picture confirms the design challenge

The World Economic Forum's Future of Jobs Report 2025, drawing on input from over 1,000 global employers representing 14 million workers across 55 economies, projects that 39% of workers' core skills will change by 2030 (World Economic Forum, 2025)down from 44% projected in 2023, but still representing the largest skills transition in the modern workforce.

The report identifies AI and data literacy as the fastest-growing skills cluster. It also identifies the most resilient roles: those involving judgment under ambiguity, cross-functional coordination, and human-centered decision-making. These are precisely the capabilities that remain in middle management after automation handles the data layer.

The automation compresses the time spent on information processing. The judgment capacity, reading a team, calibrating performance conversations, deciding what to escalate and what to resolve locally, remains. Its relative weight in the role increases. The organizational design challenge is to make that explicit: written job descriptions, clear decision rights, defined escalation logic.

The redesign advantage

Organizations that have moved to define what middle management means in an AI-augmented environment are not simply managing a transition more smoothly. They are building a management layer that functions as an amplifier: translating AI-generated signals into organizational action, calibrating team performance against real-time data, and making the judgment calls that remain outside the scope of automated systems.

The WEF 2025 report projects net job creation through 2030, 170 million new roles against 92 million displaced. The management roles that survive and expand in that environment will be the ones whose mandate was redesigned before the automation arrived, not in response to it.

The org chart changes when the decision map does.

Sources: McKinsey & Company, Middle managers hold the key to unlock generative AI (2024) · Gartner, employee survey, October 2024 (n=3,500) · Gartner, CxO survey, December 2025 (n=197) · World Economic Forum, Future of Jobs Report 2025