Copy article

Leading AI expert says "too many UK businesses are treating AI like a medieval treadmill"

ended 22. September 2025

ONE of the UK's leading AI experts and authors has warned that "too many UK businesses are treating AI like a medieval treadmill, something to make hauling heavy stone up the cathedral faster, not reduce the overall burden on their staff". Another urged companies to “design AI with strategy, ethics and responsibility at the very core so efficiency empowers people rather than exploit them”.

Meanwhile, an HR warned that the relentless pressure from some employers to maximize value from AI “could result in burnout, at scale”.

The warnings follow new research from Quantum Workplace tracking over 700,000 workers revealing that 45% of frequent AI users report high burnout compared to 35% of non-users. Translate: the very tools being marketed as productivity saviours are making things worse for the average employee, not better.

The experts say the problem is that most organisations treat AI as a capacity multiplier rather than a workload reducer. In short, if someone completes tasks 30% faster with AI, management doesn't let you leave 30% earlier—they give you 30% more work.

Colette Mason, Founder & AI Systems Architect at Clever Clogs AI, said businesses should be using AI to reduce the daily grind and enable employees to deliver better and fresher thinking: “The brutal truth around AI and real-world business? Too many UK businesses are treating AI like a medieval treadmill, something to make hauling heavy stone up the cathedral faster, not reduce the overall burden on their staff."

She added: “When companies deploy these tools without redesigning workflows, they add cognitive overhead — forcing their employees to learn new systems, check AI outputs and fix errors — while maintaining the same relentless pace. It's automation theatre: the appearance of efficiency masking deeper dysfunction.”

Mason continued: "Real AI support must lead to less hourly grind, freeing up fresher minds for strategic thinking and innovation, not faster draconian hamster wheels.”

Mitali Deypurkaystha, CEO at Newcastle upon Tyne-based Impact Icon AI, said AI isn't the problem per se, but rather management teams wanting to wring maximum value out of its adoption at their staff's expense: “This is what happens when fools rush in. AI isn’t the villain—short-sighted adoption from FOMO-driven business leaders is.

“A recent report from the Massachusetts Institute of Technology (MIT) found that 95% of generative AI pilots stall, failing to deliver meaningful ROI. Is it any wonder these leaders heap more work on employees in a scramble to justify investment?” 

Deypurkaystha continued: “The problem is that employees are paying the price for leaders' missteps. History echoes how technological advancements are touted as 'good for all' but then benefit the elite few: the Industrial Revolution boosted factory-owners while workers endured exhaustion; the internet promised inclusion yet in some cases widened divides. Now AI risks repeating the cycle—burnout for the many, profit for the few.

"We must do better: design AI with strategy, ethics and responsibility at the very core so efficiency empowers people, not exploits them.”

It's a view shared by Kate Underwood, Managing Director at Southampton-based Kate Underwood HR and Training: "AI was supposed to streamline systems but the worry is that it will create more employee stress. Improved productivity is the theory, but the reality risks being increased employee pressure as businesses seek to wring every last piece of benefit out of it.

"If AI enables you to complete your work 20% quicker thanks to AI, that shouldn’t mean your boss piles on an additional 20% of workload.

“AI should be the great enabler, streamlining core processes and enabling staff to think more freely. Instead, the relentless pressure from some employers to maximize value from it could result in burnout, at scale.”

7 responses from the Newspage community

Copy all

Star Quote
Copy

This is what happens when fools rush in. AI isn’t the villain—short-sighted adoption from FOMO-driven business leaders is. A recent report from the Massachusetts Institute of Technology (MIT) found that 95% of generative AI pilots stall, failing to deliver meaningful ROI. Is it any wonder these leaders heap more work on employees in a scramble to justify investment? The problem is that employees are paying the price for leaders' missteps. History echoes how technological advancements are touted as 'good for all' but then benefit the elite few: the Industrial Revolution boosted factory-owners while workers endured exhaustion; the internet promised inclusion yet in some cases widened divides. Now AI risks repeating the cycle—burnout for the many, profit for the few. We must do better: design AI with strategy, ethics and responsibility at the very core so efficiency empowers people, not exploits them.
Star Quote
Copy

The brutal truth around AI and real-world business? Too many companies are treating AI like a medieval treadmill, something to make hauling heavy stone up the cathedral faster, not reduce the overall burden on their staff. When companies deploy these tools without redesigning workflows, they add cognitive overhead — forcing their employees to learn new systems, check AI outputs and fix errors — while maintaining the same relentless pace. It's automation theatre: the appearance of efficiency masking deeper dysfunction. This isn't collaboration, it's digital babysitting. Real AI support must lead to less hourly grind, freeing up fresher minds for strategic thinking and innovation, not faster draconian hamster wheels.
Copy

The research might show AI users report higher burnout, but correlation doesn’t equal causation—businesses with stressed workers may be more likely to turn to AI, and without it their burnout could be worse. Let’s be honest: wages and productivity have both stagnated, so to continue to increase our standard of living we all need to do more for less. That means the time saved by AI almost inevitably gets filled with more tasks, or at least higher-value ones. All businesses will need to work even smarter moving forward, and the key is whether leaders use AI to enhance work, wellbeing, and ultimately wages, rather than simply accelerating the treadmill.
Copy

The current approach optimizes for machine efficiency, not human sustainability, as companies treat AI as a capacity multiplier rather than a workload reducer. AI deployment without workflow redesign adds cognitive overhead, learning systems, checking outputs, fixing errors - while maintaining the relentless pace. The fix requires measuring human outcomes, not just throughput, redesigning jobs, not just tasks, and giving workers agency in how AI reshapes their roles. The line between empowerment and exploitation sits at worker control, when AI gives people more autonomy versus becoming another layer of imposed surveillance and acceleration.
Copy

AI was supposed to streamline systems but the worry is that it will create more employee stress. Improved productivity is the theory, but the reality risks being increased employee pressure as businesses seek to wring every last piece of benefit out of it. If AI enables you to complete your work 20% quicker thanks to AI, that shouldn’t mean your boss piles on an additional 20% of workload. AI should be the great enabler, streamlining core processes and enabling staff to think more freely. Instead, the relentless pressure from some employers to maximize value from it could result in burnout, at scale.
Copy

AI is a productivity booster when used correctly. In a competitive labour market, it’s normal for employer expectations to adapt as better tools become available. Those who embrace AI will outperform colleagues who are slower to adapt, and may find better opportunities open up as a result.
Copy

AI has arrived with an almighty bang, disrupting countless industries. We’ve already had clients asking to shift from using human copywriters to ChatGPT, expecting reduced costs or increased output, based on the misguided belief that AI can deliver the same quality for less. The intelligence in emerging learning systems like ChatGPT are not always accurate and AI chat results are notoriously easy to outsmart. AI agents never say to you, ‘I don’t know’, but will press on, giving you answers with absolute confidence. Anyone with mastery of a subject will have found that conversational topics can see AI asserting ‘facts’ that aren’t accurate, and when challenged your chat agent will acquiesce to your correction and alter its answer. Working with a low knowledge base sycophant, even one as fast and tireless as AI, is not helpful when it comes to getting the job done properly.