When Artificial Intelligence first entered the mainstream, many leaders expected the usual promise: faster execution, lighter workloads and a welcome reduction in the daily clutter of work. Instead, quite a few managing directors and senior teams are finding themselves in a different position entirely. They are producing more drafts, more documents and more output than ever, yet somehow feeling busier, not freer.
That is the productivity paradox.
AI increases workload when teams use it for one-shot finished outputs rather than structured components. EY’s 2025 Work Reimagined Survey of 15,000 employees across 29 countries found 64% reported increased workloads, while only 5% used AI in advanced, transformative ways. The productivity gain comes from workflow design — breaking work into managed components and defining standards upfront — not from the tool itself.
AI is meant to save time, yet many firms are spending hours rewriting bloated drafts, cleaning up weak outputs and sorting through machine-generated noise that never should have made it into the workflow in the first place. The result is frustration, scepticism and the creeping suspicion that the technology may be adding effort rather than removing it.
“AI does not save time simply because you switched it on. It saves time after you have done the thinking, the setup, and the workflow design.”— Larysa Hale, Expert Circle
In many cases, that suspicion is understandable. But the real problem is not AI itself. It is the way people are trying to use it.
Why does AI increase workload instead of reducing it?
A great deal of disappointment with AI comes from what I would call the one-shot trap. Leaders and teams ask for a finished masterpiece in a single prompt, receive something broad or off-point, and then spend more time correcting it than they would have spent doing the work properly from the start.
That is not a productivity strategy. That is a badly managed shortcut.
It is a bit like buying a high-end television, plugging it in, and complaining that it does not work because you never configured the channels, connected the devices or sorted the Wi-Fi. The technology may be excellent, but until it has been set up for the reality of your environment, it will not perform as expected.
AI works much the same way. It is not a shortcut around work. It is a force multiplier for well-structured work.
If you want it to save time, you have to invest time first. You need clarity on the task, the desired outcome, the standard, the data, the tone and the business purpose behind the output. Without that setup, AI generates volume rather than value.
What the evidence says about AI and productivity
The data suggests this is not just a handful of firms getting it wrong. It is a broader pattern.
EY’s 2025 Work Reimagined Survey, which surveyed 15,000 employees and 1,500 employers across 29 countries, found that 88% of employees use AI at work, but mostly for basic tasks such as search and summarisation. At the same time, 64% reported a perceived increase in workload, while only 5% said they were using AI in more advanced, transformative ways. EY’s conclusion was blunt: many organisations are still missing a substantial share of the productivity gains AI could deliver because adoption has outpaced capability and workflow design.
MIT Sloan has made a similar argument: leaders get better outcomes when they redesign work, deconstruct tasks and rebuild workflows around AI, rather than simply dropping a tool into an unchanged process and hoping for magic.
“Too many teams use AI to generate volume, then wonder why they are drowning in edits. Productivity comes from structure, not from speed alone.”— Larysa Hale, Expert Circle
How to make AI save time instead of adding work: three changes
The first shift is to stop asking AI for the whole pie. Break work into smaller, managed deliverables. Ask for the hooks, then the structure, then the synthesis of evidence, then the first draft. That sounds less glamorous than a miracle prompt, but it is much more commercially reliable.
The second shift is to invest in system setup. If your team is repeatedly using AI for marketing, proposals, analysis or internal communication, define the standards once. Establish the tone, the approved source material, the commercial goals and the quality criteria. That upfront investment saves endless rewriting later.
The third shift is to automate the process, not merely the output. Leaders often become fascinated by the draft AI produces, when the real opportunity is in building a repeatable workflow around the repetitive parts of the job. The productivity gain comes from the operating model, not from the novelty of the generated text.
If AI is making your team busier, the question is not whether the technology has failed. The question is whether your firm is using it in a disciplined enough way to create value instead of noise.
Speed without standards is not productivity. It is just a faster route to more work.
Our briefings and programme are designed for senior leaders who want to move beyond shallow use and turn AI into a genuine force multiplier for the business. Because real productivity does not come from asking AI to do everything. It comes from knowing exactly what part of the work it should do, how it should do it and what standard it still has to meet.
Larysa Hale is the founder of Expert Circle and creator of the AI-Driven Marketing Growth Programme, a structured series of briefings and masterclasses for managing directors and senior leaders in professional services. She has spent over 15 years helping founders, marketing directors and business leaders build commercially grounded growth strategies.
