The myth and the reality

Middle managers have always been anxiety-prone during transitions. In the 1990s, it was the internet. In the 2010s, it was automation and outsourcing. Now it's AI. And there's a reason the anxiety keeps returning: middle management has genuinely been the easiest layer to flatten when technology makes it possible.

But here's what the data actually shows in 2026: middle management positions haven't collapsed. What's changing is the composition of the role. AI is automating the administrative overhead — scheduling, status updates, routine reporting, meeting organisation — while the strategic and interpersonal work remains stubbornly human. The job is shedding fat, not disappearing. Recent research from McKinsey confirms this pattern across industries: the managers thriving are those shifting focus from process management to talent development.

The real risk isn't redundancy. It's skill mismatch. Managers trained to excel at administration and process management are finding those skills suddenly less valuable. Managers trained to coach, facilitate cross-team problem-solving, and make judgment calls under uncertainty are finding themselves more essential than ever. Harvard Business School research on management transitions documents this pattern: roles that emphasise technical execution are losing ground to those centred on people development.

The shape of the role is changing, not the role itself. What's being automated is the stuff nobody enjoyed anyway — the busywork. What remains is the skilled, human-centric work that's actually hard to do well.

What's being automated, and what isn't

To understand where middle managers are affected, it helps to separate management functions into the automatable and the irreducibly human.

Management Function What's Changing What Still Requires a Human
Status updates & reporting AI compiles data, generates dashboards, summarises team progress Interpreting what the data means for strategy; having honest conversations about blockers
Meeting scheduling & coordination Calendar AI books meetings, finds overlaps, sends reminders Deciding whether a meeting is necessary; reading the room during difficult conversations
Routine decision documentation AI logs decisions, drafts meeting notes, creates knowledge base entries Making judgment calls when information is incomplete; weighing intangible factors
Performance data collection AI gathers metrics, flags anomalies, prepares feedback frameworks Coaching conversations; delivering tough feedback with empathy; developing growth plans
Onboarding workflows AI delivers training content, tracks completion, sends reminders Welcoming new hires; assessing cultural fit; mentoring high-potential people
Project timeline tracking AI monitors schedules, flags delays, suggests resource reallocation Negotiating trade-offs; unblocking people stuck on hard problems; building trust with stakeholders

The pattern is clear: information processing and coordination is being handed to machines. Judgment, empathy, and persuasion remain stubbornly human. PWC's 2026 AI report shows that 68% of organisations implementing AI in management workflows report improved manager effectiveness — but only when paired with upskilling in soft skills.

A typical manager's week: before and after

Concrete example: meet Sarah, a software engineering manager at a mid-size tech company. Here's how her typical week was structured in 2023, versus 2026.

2023 Typical Week 2026 Typical Week
Monday morning (1 hour): Compile status updates from team Slack into a report for the director Monday morning (10 min): Review AI-generated team summary; use freed time to prep 1:1 agenda
Monday–Friday (3–4 hours/week): Back-and-forth emails about meeting times; calendar shuffling Monday–Friday (0 hours): Calendar AI handles all scheduling
Wednesday (2 hours): Prepare and run team standup; manually consolidate blockers Wednesday (45 min): Facilitate deeper discussion on two critical blockers that AI flagged as risky
Friday (1.5 hours): Write performance notes for three team members; log feedback into HR system Friday (2 hours): Have coaching conversations with three people; AI transcribes and summarises for record
Ad hoc (2–3 hours/week): Chase down project status from different teams; create meeting notes Ad hoc (0 hours): AI keeps shared knowledge base current; notifies only if consensus is needed
High-value activity (2–3 hours/week): 1:1 coaching, cross-team problem-solving, strategic planning High-value activity (8–10 hours/week): Deep coaching, mentorship, unblocking complex problems, strategy

Sarah's workweek isn't shorter — her company hasn't reduced her hours. But her discretionary time has quadrupled. The busywork that consumed 60% of her week has been automated. The question isn't whether she still has a job. It's whether she can learn to fill that time with the kind of strategic, developmental work that AI can't do, and that organisations actually value more. LinkedIn's 2026 workforce report notes that managers who successfully transitioned to high-discretionary roles have seen average engagement scores from direct reports rise 34%, and retention improve measurably.

Who's at risk, and what happens next

Not all middle managers are equally vulnerable to this shift. The risk profile depends on where the role sits and what it emphasises.

Highest risk: Managers whose value comes primarily from information synthesis and process execution. Think project managers who spend 80% of their time scheduling, tracking Gantt charts, and producing status reports. These roles are most directly displaced by AI tooling.

Moderate risk: Managers in fast-scaling organisations where administrative burden has been high, and who haven't yet built coaching or strategic skills. They need to upskill quickly or find themselves sidelined even if the position isn't eliminated.

Lowest risk: Managers known for developing talent, navigating cross-functional complexity, and making decisions under uncertainty. These skills — which AI doesn't automate — are becoming rarer and more valuable.

What's urgent is retraining. For managers, this means:

For organisations rolling out AI tooling, the shift is already happening whether you're intentional about it or not. Being deliberate helps. Audit your managers to understand who was thriving on execution versus who was already leaning into coaching. Invest in development deliberately — pair automation rollout with coaching and strategy training, so people understand how to spend freed-up time. Expand how you measure manager success: stop rewarding reports filed or meetings scheduled; start measuring coaching impact, cross-team unblocking, and long-term capability building. And be honest about who won't make the transition: not every manager will become a skilled coach, and that's okay. Some will be better suited to individual contributor roles or different teams. Better to help them find the right fit than to watch them struggle and disengage.