AI Won't Replace Managers — But Managers Who Use AI Will

Here's the data point that should keep every manager up at night. According to Gallup's Q4 2025 survey, 69% of organisational leaders now use AI at work at least a few times a year. For managers, that figure drops to 55%. For individual contributors, it's just 40%. The people at the top are adopting AI fastest — and they're using it to make decisions about the people below them.

This isn't a hypothetical future. Microsoft's AI Economy Institute reported in early 2026 that global AI usage has reached record levels, but with a widening divide between organisations that have integrated AI into their workflows and those still figuring out where to start. In technology, 77% of workers now use AI. In finance and professional services, adoption is surging. In retail, it's flat at 33%.

The split isn't between industries. It's between managers who treat AI as a tool and those who treat it as a threat.

What AI Actually Does to Management Work

A recent MIT Sloan study found that access to AI tools increased productivity for knowledge workers by over 40% — largely by automating information synthesis and retrieval. These are the tasks that have historically consumed a manager's day: compiling status reports, summarising meetings, cross-referencing project data, and preparing for reviews.

Let's be specific about what's changing right now.

1-on-1s. Tools like Fireflies.ai and Otter.ai now transcribe, summarise, and extract action items from meetings automatically. Notion AI drafts follow-up summaries and surfaces overdue tasks from previous conversations. The manager who used to spend 30 minutes preparing for every 1-on-1 and 15 minutes writing up notes afterwards can now redirect that time to the actual conversation.

Project coordination. Asana Intelligence and ClickUp Brain automatically identify at-risk tasks, suggest timeline adjustments, and flag resource bottlenecks. Microsoft 365 Copilot, since late 2025, can build custom workflows from natural language — a manager can describe a process and have Copilot generate the automation. This is the coordination work that Harvard Business School research identified as the primary candidate for AI-driven flattening of corporate hierarchies.

Decision support. IBM reports that AI-powered project management tools now summarise documentation, draft stakeholder updates, and surface answers from large knowledge repositories — functions that reduce communication overhead and improve clarity. McKinsey estimates AI management tools can reduce management overhead costs by 15–40%, depending on the industry.

Hiring and onboarding. AI-assisted screening, interview transcription, and onboarding workflow automation are becoming standard. The manager's role shifts from processing candidates to making judgment calls about culture fit and potential.

The Manager Who Refuses to Adapt

There's a category of manager that AI is genuinely coming for: the one whose primary value lies in holding information, producing reports, and resolving routine problems. The Chartered Management Institute estimates that 82% of managers in the U.K. receive no formal preparation for the people-management aspects of their role. They were promoted for technical competence, not leadership ability. Their leverage came from being the person who knew things.

Knowledge is no longer power. Knowledge is ubiquitous. And the manager whose job description could be summarised as "human database with scheduling privileges" is in trouble.

The Observer's analysis of this shift is blunt: "When the 'what' and the 'how' of a task are automated, what is left for a manager to do?" The answer is everything that truly matters: the "who" and the "why."

What AI Cannot Do (Yet)

This is where nuance matters, because the AI-will-replace-everyone narrative is as reductive as the AI-will-change-nothing counter-narrative.

AI cannot build trust. It cannot sense that a high-performing team member is burning out before they quit. It cannot navigate the politics of a cross-functional project where three VPs have conflicting priorities. It cannot have the difficult conversation about performance that everyone is avoiding. It cannot create psychological safety — the condition that Google's Project Aristotle identified as the single most important factor in high-performing teams.

Gallup's global data reveals the scale of the challenge: only 10% of U.K. workers feel engaged at work. Globally, 41% of employees experience high daily stress, rising to nearly 60% for those under poor management. Disengagement and burnout cost the global economy an estimated $8.9 trillion annually — roughly 9% of global GDP.

No AI can fix this. Only managers can. But managers who are freed from administrative overhead by AI have more capacity to try.

The Practical Playbook

So what does a manager who uses AI actually do differently? Based on the tools available today and the organisations seeing results, the pattern is consistent:

Audit your time ruthlessly. For one week, log every task you do. Categorise it: information synthesis, coordination, communication, decision-making, relationship-building. The first three categories are increasingly automatable. If they dominate your schedule, start there.

Layer AI into existing workflows, don't add new ones. The biggest failure mode is adopting AI tools as bolt-ons that create more work. Use Copilot inside the Teams meetings you're already having. Use Notion AI within the project wiki your team already uses. Use Fireflies on the Zoom calls that are already on your calendar.

Reinvest the time. This is the step most managers skip. If AI saves you 45 minutes a day on report-writing and meeting prep, and you fill that time with more email, you've gained nothing. Redirect it toward the work that only you can do: coaching, stakeholder alignment, strategic thinking, and the human conversations that determine whether people stay or leave.

Build AI literacy in your team. The Gallup data shows a clear hierarchy: leaders use AI more than managers, who use it more than individual contributors. This gap is a choice. Managers who actively teach their teams to use AI tools multiply their impact. Those who don't risk creating a two-tier team.

The Real Competitive Advantage

IESE Business School's 2025 research on AI and management found that AI tools empowered employees with real-time information, enabling quicker decisions and reducing reliance on managers' judgment. This decentralisation made companies flatter and more agile. But it also meant fewer managers overseeing larger teams — and each of those managers needed stronger human leadership skills, not weaker ones.

The most in-demand managers of 2026 are not the most technical ones. They're the ones who can combine AI-driven efficiency with genuine human connection. They run meetings that matter, not meetings that could have been an email — or a Copilot summary. They make their teams faster and more cohesive. They use AI to scale their reach without scaling their distance.

The managers who refuse to engage with AI won't be replaced by algorithms. They'll be outperformed — quietly, consistently, and at scale — by peers who figured out what to automate and what to protect.

The trapdoor isn't beneath every manager. It's beneath every manager who mistakes activity for impact, and information for leadership. AI didn't build that trapdoor. It just made it visible.