AI Is Making Us More Productive.

So Why Are We Still So Busy?

Artificial intelligence was supposed to save us time.

That has been one of the biggest promises attached to workplace AI: fewer repetitive tasks, faster admin, smarter workflows, more room for the work that actually requires a human brain.

And according to new research, employees do feel more productive because of it.

Culture Amp’s latest AI at Work benchmark found that 71% of employees say AI helps them feel more productive, rising to 93% among the heaviest users.

So far, so good. But then comes the slightly awkward bit. Workloads do not appear to be falling.

Among AI power users, 72% say their workload feels reasonable for their role. Among people who do not regularly use AI, that figure is 69%. In other words, despite a very large gap in perceived productivity, the difference in how manageable work feels is tiny.

And that raises a much bigger question.

If AI is saving us time, where is that time actually going?

The productivity paradox

This is the bit businesses need to pay attention to.

If somebody can now complete a five-hour task in three hours, that is a genuine productivity gain.

But what happens to the two hours saved?

Do they become time for better thinking?

• More creativity?

• Training?

• Development?

• A calmer workload?

• A proper lunch break?

• Or do they simply become room for another five tasks?

Culture Amp’s research asks exactly this. The study points out that businesses now need to decide what the “AI dividend” is actually for: more output, better-quality work, reduced pressure, innovation, development or some combination of these. Because if every hour saved by technology immediately becomes another hour filled with work, employees may become more productive without ever feeling less stretched.

That is not necessarily transformation. It may simply be acceleration.

Faster does not always mean better

This is where workplace conversations around AI can become a little too simplistic.

We ask:

• How many people are using it?

• How much time is it saving?

• How many tasks can now be completed faster?

All useful questions.

But they are not the whole story.

An organisation could have very high AI adoption and still have employees who are overloaded, unclear about what the technology is actually for, worried about their future career paths and spending time correcting poor-quality AI output.

That last point matters too.

Separate research referenced in the article found that some employees spend several hours each week correcting poor-quality AI-generated work.

So AI can save time in one part of a workflow while quietly creating work somewhere else.

Anyone who has ever received a beautifully formatted AI-generated document containing one completely invented fact will understand the issue.

Efficiency is only useful if the work is still good.

Employees seem ready. Leadership may be playing catch-up.

One of the most interesting parts of the Culture Amp findings is that employees do not appear particularly resistant to AI. Quite the opposite. Some 85% say their organisation encourages AI experimentation.

But only 58% say leaders have clearly explained where their organisation is actually going with the technology. Only 60% say their manager shares practical examples of how AI can support their work. That is quite a gap.

Employees are effectively being told: “Use AI. Experiment. Explore.” And many are.

But the answer to the next question - why? - is often much less clear.

That matters because technology without direction can easily become another workplace expectation.

• Another tool to master.

• Another platform to learn.

• Another thing employees are supposed to somehow squeeze into an already busy day.

The interesting challenge now may not be persuading people to use AI.

It may be helping organisations work out what they actually want AI to achieve.

The best AI users may have something to teach us

There is another interesting finding in the data.

Some 72% of AI power users said their organisation motivates them to go beyond what they would in a similar role elsewhere, compared with 57% of non-users.

That does not mean AI automatically makes people more motivated. The research itself is careful not to claim that. It could simply be that more engaged employees are more likely to experiment with AI in the first place. But it does suggest there may be value in studying the people who are using these tools most effectively.

• What tasks are they delegating to AI?

• What are they still choosing to do themselves?

• Where is it genuinely saving time?

• Where does it create more work?

• What support have they had?

That feels far more useful than simply chasing adoption percentages.

The question becomes less:

“How do we get everyone using AI?”

and more:

“What are our best users doing that actually improves their working day?”

That is a much better conversation.

Then there is the career question

AI is not simply changing tasks.

It is starting to change how people think about careers.

Culture Amp found that employees’ awareness of internal career opportunities has fallen by 10 percentage points since July 2025.

That makes sense.

People are trying to work out which roles will change, which skills will matter, what AI will automate and what their job might look like in a few years. And organisations do not have all of those answers yet.

That is ok.

But silence is probably not.

The article makes a particularly useful point here: employees may not need certainty from leaders right now, but they do need honesty and ongoing communication - That feels important.

Because uncertainty filled with information is manageable.

Uncertainty filled with silence tends to become anxiety.

What should we actually do with the time AI gives us back?

This is the question we keep coming back to.

And it feels very MOCO.

Because we talk constantly about capacity.

We all have a finite amount of it.

• Time.

• Energy.

• Attention.

• Decision-making.

• Emotional bandwidth.

If AI genuinely gives some of that capacity back, the obvious temptation is to fill it immediately.

Humans are very good at doing this.

• Finish something early? Excellent. Here's something else.

• Clear half the inbox? Brilliant. Another twenty emails arrive.

• Become quicker at your job? Lovely. Your new baseline is now “quicker”.

Eventually the productivity gain disappears into expectation.

And nobody can remember when the workload increased.

That is the danger of treating efficiency as though its only purpose is more output.

Perhaps the real opportunity is better work

There is another version of this story.

One where AI removes repetitive admin and employees spend more time thinking.

• Where managers get some space back to actually manage people.

• Where someone can spend an hour learning rather than processing.

• Where teams have more time for collaboration.

• Where workloads become slightly less frantic.

• Where people leave work with enough brain power left to have a life.

These are all still a return on investment. Possibly a very good one.

The research argues that the AI dividend could include better quality, reduced administrative burden, innovation, learning, stronger skills and healthier workloads - not simply a higher volume of tasks.

And that feels like a conversation worth having.

We should be protecting some of the human space

One of the stranger possibilities of workplace AI is that we could automate repetitive work only to fill the resulting time with more repetitive work.

Which would be quite an achievement.

Instead, perhaps some of the saved capacity should be deliberately protected.

• For learning.

• Critical thinking.

• Creativity.

• Conversation.

• Career development.

• Problem-solving.

The bits of work that benefit from judgement, empathy and context.

The article specifically recommends protecting time for human development as AI takes on more routine tasks. That feels increasingly important. Because if technology gets better at doing the mechanical parts of work, then the human parts become more valuable, not less.

Managers have a huge role here

This will not solve itself through an AI policy.

Managers need to understand what their teams are using.

• Where it is working.

• Where it is creating problems.

And crucially, whether time saved is quietly becoming workload creep.

That does not mean every AI-generated efficiency should become an extra coffee break.

Businesses are obviously entitled to expect commercial value from significant investment in technology. But sustainable productivity probably requires a balance.

• Some gains may become output.

• Some may improve quality.

• Some may reduce pressure.

• Some may create development time.

The important thing is that organisations make those choices deliberately rather than allowing every efficiency gain to automatically become the new minimum expectation.

Because “more productive” should eventually feel like something

This is perhaps the simplest takeaway.

If seven in ten employees believe AI is making them more productive, but their working lives feel broadly as busy as before, then businesses should be curious about that gap.

Not suspicious. Curious.

• Where did the time go?

• Who benefited?

• What improved?

• What got added?

• Did quality change?

• Did pressure change?

• Did employees learn anything?

• Did customers get a better experience?

• Or did everyone simply start running slightly faster?

Those are the questions that will matter as AI becomes increasingly ordinary at work.

Because success should not just mean employees can do more. It should mean work itself becomes better in some meaningful way.

• Better quality.

• Better decisions.

• Better use of people's time.

• Better opportunities to develop.

Perhaps even occasionally a little less pressure.

AI may well be able to give us time back.

The next challenge is deciding whether we are brave enough not to immediately fill every minute of it.

Want to dig a little deeper?
This piece was inspired by Culture Amp’s latest AI at Work Benchmark, which explores how employees are actually experiencing AI in the workplace - from productivity and workload to leadership, career development and trust. If you’re interested in the data behind the discussion, you can read the full findings directly from Culture Amp.

At MOCO, we’re especially interested in what happens next: not just whether AI helps us work faster, but whether it helps us work better. Because if technology gives us time back, the real opportunity is deciding what we do with it.