Artificial Intelligence
The Hidden Cost of Disconnected AI
Six Thousand Employees. Four Markets.
Michael Brenner
Vice President, Thought Leadership and Customer Advocacy
Workday
One in four UK employees loses at least one full day each week to copying, pasting and translating AI’s work between tools and departments, the biggest time loss among 29 countries studied in Workday’s research into AI gains globally.
The global average is one in five employees who lost a day per week. Canadians lose less time copying and pasting, (only 12% lose more than seven hours a week), but they report being more stressed about broken processes than employees in other countries.
The Copy/ Paste Economy: How Task-Based AI is Failing the Enterprise, shows that, among workers using AI, no country is immune to the busywork that comes with systems that don't work together, especially when AI is not fully embedded into the software that runs the business.
Workday surveyed 6,100 professionals across HR, finance, IT, and operations in organizations of 500 employees or more, all active AI users, spanning North America, Europe, Asia Pacific, and beyond. The analysis revealed a consistent trend of inefficiency caused by AI deployed around the edges of the business.
Capable People, Failing Architecture
Ask employees in any of the markets we studied how they feel about their work and their answers are strikingly consistent. In the US, 97.8% of professionals rate their day-to-day work experience positively. In the UK, 97.3%. In Canada, 96.2%. In Australia and New Zealand, 96%. These are engaged, capable people who believe in what they do and believe AI can make it better.
They are also spending significant portions of every week doing the job that their systems should own. Moving data from one platform to another. Coordinating approvals through email chains because two systems cannot route them automatically. Reconciling conflicting reports because different tools hold different versions of the same number.
More than 8 in 10 employees in every region report spending significant time on this kind of coordination work. They are acting as human middleware in a copy/paste economy, moving data from one system to the next. Not because they want to. Because the architecture requires it.
That is the hidden tax.
What the Regional Data Shows
The numbers are not identical across markets. In the United States, only 26% of organizations have moved AI from the periphery into core workflows. One in five employees loses more than seven hours a week to moving information and reconciling data between systems that cannot do it automatically..
Report
1 in 5 US employees loses more than 7 hours a week to the hidden tax on your workforce.
In the UK, one in four employees loses a full day a week to coordination work, worse than the global average of one in five. Only 22.5% of UK organizations have AI working inside their core systems, the lowest of any region in this study.
And 59.5% of UK employees say they often or very often end the day feeling busy but not genuinely productive. That is the highest reading of any market we measured. The hidden tax in the UK is not just on the clock. It is about confidence.
Report
In the UK, 1 in 4 employees loses more than 7 hours a week. And fewer organizations have AI in core systems than anywhere else we looked.
The results from Canada tell a different story. The time lost is lower: 12.5% of Canadian employees lose seven or more hours a week, well below other markets. And 65% say AI has already cut their task completion time, the highest rate of any region we studied.
By those measures, Canada is ahead. But process stress runs at 85%, the highest of all four markets. The hidden tax in Canada is not showing up in lost hours. It shows how people feel while they work.
Report
65% of Canadian employees say AI has cut their task completion time. Process stress is still the highest of any market.
And then Australia and New Zealand changed the argument entirely.
ANZ organizations have embedded AI into their core systems at a higher rate than any region in this study: 30.8%. More AI in more systems than the US, the UK, or Canada. And yet 89% of ANZ employees still spend significant time moving information between systems, the highest of any region we studied.
Thirty percent lose more than seven hours a week to reconciling conflicting data alone, higher than the global average.
Report
In ANZ, 30% of employees lose more than 7 hours a week to reconciling conflicting data alone. More AI in core systems than any region. The glue is still human.
When More AI Isn't Enough
The Australia and New Zealand data is notable because the region has the highest AI integration of any region in this study. Workers in this region have more tools. But they also report spending significant time coordinating between systems. The tax is not lower. It is at least as high.
That is because the question was never how much AI you have. It was where that AI lives in your operating model.
When AI sits alongside the workflow, as a separate tool whose output an employee carries manually to the next system, you get individual efficiency gains. The employee works faster. Then they paste the result somewhere. Then they wait for someone to check it. The coordination work remains.
When AI runs inside the workflow, handling the monitoring, the routing, the reconciliation, the approval chains that currently run through human inboxes, the coordination work moves from the employee to the platform. That is where the tax gets paid down. Not by adding AI. By changing where AI operates.
A Problem Without Borders Needs a Solution Without Borders
The regional data gives leaders in global organizations something more useful than a benchmark. It offers a diagnostic.
In markets where AI integration is lowest, the priority is architectural: the operating environment has to change before AI can work at scale. In markets where integration is higher but the coordination burden persists, the question is not whether AI is present but whether it is inside the workflow or sitting beside it. In markets where time savings are visible but stress remains high, the systems have changed faster than the processes built around them.
The prescription is the same everywhere. Where you start depends on where the friction is worst.
What does not change by region is what employees are saying. They like their work. They believe AI can improve it. They are already using AI to patch the holes the architecture created, because nobody has fixed the architecture yet.
The good news is that this global problem has a global solution. It lives within the platform.
The Copy/Paste Economy
More than 8 in 10 employees in every region report spending significant time acting as human middleware.