I learned early in my consulting career that clients would pay for almost anything except thinking.

They'd sign off on deliverables, decks, workshops, and implementation support without blinking. But time spent figuring out what actually mattered — framing the problem, weighing trade-offs, deciding what not to do — that was harder to sell. Thinking was tolerated only insofar as it was invisible, bundled into something tangible that looked like work.

Inside companies, the same dynamic persists. Managers are comfortable managing what they can see: activity, process compliance, responsiveness, hours logged. What they struggle to manage is judgement — the slow, non-linear work that doesn't photograph well but determines whether anything else matters.

This is why the latest workplace conversation shouldn't surprise anyone.

The task-masking moment

"Task-masking" is the latest label for employees appearing busy without doing meaningful work. Charging around with a laptop. Pretending to be on urgent calls. Typing furiously with no real purpose. The behaviour isn't new — faking productivity is an age-old skill, practised and perfected long before TikTok existed. What's new is that people are now openly discussing it, even filming themselves doing it.

The timing isn't coincidental. Task-masking videos started gaining traction as major employers enforced stricter return-to-office mandates. When companies signal that presence equals productivity, employees optimise for presence. The behaviour that was always there became visible — and, for a generation used to documenting everything, shareable.

Most commentary treats this as a problem to be solved. Clearer values. Broken-down projects. Psychological safety. The advice isn't wrong, exactly. It's just shallow.

Task-masking is not a workforce failure. It's a management failure made visible. And AI doesn't solve it — it removes the last places it can hide.

When output becomes illegible

Modern knowledge work produces value in ways that are hard to observe. Insight rather than artefacts. Avoided mistakes rather than visible wins. Decisions rather than documents.

The relationship between effort and output has weakened. A good decision might take an hour and save a year. A bad one might take weeks and look productive all the way through.

In that environment, organisations substitute signals for substance. Busyness becomes a proxy for value. Visibility replaces impact. Responsiveness stands in for progress.

This is where task-masking comes from. Not laziness — but rational behaviour in a system that rewards appearance over outcomes. When you're measured on looking busy, you optimise for looking busy. The workforce isn't broken. The measurement is.

The management gap underneath

Here's the uncomfortable context: 82% of UK managers become managers without any formal management and leadership training. The Chartered Management Institute calls them "accidental managers" — people promoted because they were good at their previous job, then left to figure out management through trial and error.

The UK has approximately 2.4 million of them.

These managers default to what they can observe: activity, presence, responsiveness. Not because they're bad people, but because no one taught them to manage anything else. They track tasks because tasks are visible. They count hours because hours are countable. They value busyness because busyness is the only signal they know how to read.

Task-masking is what happens when a workforce adapts to management that can only see motion.

The org chart problem

Traditional management structures make this worse. I've written before about how org charts show where people sit, not where decisions get made. They describe reporting lines, not judgement. They're good at managing execution and poor at managing capability.

When all work was done by humans, these views aligned closely enough that the difference barely mattered. If you knew who owned marketing, you had a reasonable mental model of how marketing decisions happened. The people and the decisions lived in the same place.

But the gap was always there. Org charts never showed how work actually flowed — where judgement was required, what happened when exceptions arose, what got decided in the spaces between boxes. They showed the formal structure but not the informal one. Anyone who has worked in a company of any size knows the actual work often happens through relationships no chart captures.

AI is about to make this gap much wider.

What AI actually changes

AI is often described as a productivity tool. That understates the impact.

Generative AI collapses the effort-output relationship. Work that once took days now takes minutes. Agentic systems execute processes perfectly and invisibly. Output floods the system.

Three long-standing management crutches disappear at once. First, visible effort — AI doesn't look busy. Second, process as reassurance — AI follows process flawlessly. Third, output as value — AI produces more output than anyone can absorb.

What remains is selection, prioritisation, judgement, and restraint. In other words: capability.

AI doesn't automate jobs first. It automates the assumptions bad management relied on.

Why weak managers struggle more

There's an assumption that AI will help managers manage. In practice, it raises the bar.

Managers who rely on task tracking, utilisation metrics, meeting density, and constant visibility find that none of these correlate with contribution anymore. The junior analyst's box on the org chart still exists, but the work inside it has transformed. The human reviews and refines; the AI drafts and processes. The org chart shows the same structure, but the capability distribution has shifted entirely.

The gap widens between those who can evaluate reasoning and those who can only supervise activity. AI raises the floor of execution but raises the ceiling of judgement even faster.

If you couldn't manage thinking before AI, you won't manage it after.

The predictable failure mode

Most organisations will respond in a familiar way: layer AI onto existing structures, demand more output faster, add more metrics to "stay in control."

This doesn't create transformation. It creates what I'd call output-masking — faster, cheaper noise that still avoids the hard questions. The AI generates the reports. The dashboards fill with metrics. The presentations multiply. Everyone looks productive.

But no one is asking whether any of it matters. The organisation has simply automated its avoidance of judgement.

Task-masking was employees appearing busy without producing value. Output-masking is organisations appearing productive without making decisions. Same dysfunction, higher volume, lower cost per unit of waste.

What survives

Management doesn't disappear in an AI world. But it changes shape.

The managers who remain useful clarify intent rather than assign tasks. They define decision boundaries rather than workflows. They evaluate judgement without hindsight bias. They treat thinking as the primary human contribution.

This is why capability-based thinking matters — it gives managers something real to manage when tasks no longer define value. Instead of asking "what does this person do?" you ask "what decisions does this role make, within what constraints, and under what conditions does it hand off?"

That question works whether the role is filled by a human, an AI, or some combination. The boundaries are what matter. The intelligence filling those boundaries can vary.

Managers who understand this have a future. Managers who are still counting keystrokes do not.

The uncomfortable conclusion

AI will not fix bad management. It will expose it — quietly, relentlessly, and without negotiation.

Organisations that continue to manage activity will drown in automated output. Those that learn to manage capability, judgement, and outcomes will find AI amplifies their effectiveness.

The task-masking conversation has it backwards. The problem isn't employees who fake productivity. It's organisations that made faking productivity the rational choice. AI doesn't solve that problem. It scales it — until the gap between motion and value becomes impossible to ignore.

The technology is not the bottleneck. Management always was.

The takeaway

If your organisation is worried about task-masking, don't start with employee behaviour. Start with what you're actually measuring and rewarding. If the answer is activity, presence, and visible busyness, you've designed a system that optimises for exactly what you're complaining about.

AI is about to make that design choice very expensive.