Behind the promise of AI efficiency, employees talk about a more nimble office culture that eliminates easy tasks and expands those that require judgment.
For years, artificial intelligence has been imagined as the ultimate workplace liberator, a tool that can eliminate repetitive tasks, shorten working hours, and give people more time to think.
But many workers say the opposite is happening, as the tools are now part of everyday office life.
Far from reducing workloads, they say, AI is accelerating the pace of work to exhausting levels.
This pressure could be particularly acute in South Korea, where an office culture of long working hours and quick responses is already deeply ingrained.
According to a report released last month by the Korea Institute for International Economic Policy, South Korea had the highest rate of AI use in the workplace among the four countries surveyed, at 51.1%, exceeding Japan and Taiwan.
“At first, AI felt like magic,” says Kim, a data engineer at a major Korean search platform company who has worked in data engineering and machine learning for more than a decade. “But now I think this is the main cause of my fatigue these days.”
“The tempo of work is incredibly fast now,” Kim said.
Tasks that once required three or four engineers to complete a week can now be completed by one worker in half a day using AI coding tools, he said. But that doesn’t mean the job is any lighter.
“Everyone is using AI now, so expectations have changed quickly,” he said. “The easy parts of the job have gone and the hardest parts have increased.”
Kim now spends less time writing code and much more time reviewing AI-generated output, coordinating with other departments, assessing service risks and validating results, which he says is mentally draining.
“Coding was often easier than communicating, judging, and validating,” he said. “These parts take up most of the day.”
When productivity is under pressure
AI tools can improve productivity. But instead of giving workers more leeway, many companies use those profits to demand more and faster production.
According to a recent study by Upwork Research Institute, 81% of managers in the US and UK say AI has increased their expectations for employee productivity. At the same time, 71 percent of employees reported burnout and 65 percent said they were struggling under increased productivity pressure.
Many employees today juggle multiple AI systems simultaneously, using one tool for research, another for document summarization, and another for drafting presentations and analyzing data.
Boston Consulting Group recently used the term “AI brain fly” to describe the cognitive overload caused by excessive interaction with multiple AI systems in a study published in Harvard Business Review.
Unlike traditional burnout, which involves emotional exhaustion and cynicism about work, brain fly is more like a mental deadlock: chronic headaches, difficulty concentrating, slow decision-making, and what some workers describe as “brain fog.”
The researchers found that productivity increased when employees used one or two AI tools, but productivity dropped sharply when employees started working with four or more systems simultaneously.
“The brain is constantly switching between situations,” says Hyun Myung-ho, a psychology professor at Chung-Ang University. “Employees are driving output nonstop, reviewing, comparing, revising, and fact-checking. That level of cognitive monitoring creates a tremendous mental strain.”
Part of the exhaustion is due to the fact that AI tends to create entirely new categories of jobs.
Lee, a marketing analyst in Seoul, said he often spends hours revising AI-generated reports that management assumes are already mostly complete.
“My boss thinks AI can do a week’s worth of work in a day,” Lee said. “But in reality, the editing effort has doubled because the output requires human judgment.”
Chong, a content creator in her 20s, said the explosion of AI-generated videos and images has also increased pressure within the creative industry.
“Now there is an endless amount of content,” she said. “Algorithms don’t recognize my work the way they used to, and I feel powerless when competing with the endless amount of AI-generated content.”
One of the biggest problems, employees say, is that AI often produces output that looks sophisticated and convincing, even when it contains errors.
Engineer Kim described the explosion of what he called “fake productivity.”
“Even people with poor performance can now generate large amounts of seemingly plausible code,” he said. “But someone still needs to verify whether any of it actually works.”
That burden often falls on senior employees. Reviewing defective work with AI assistance, identifying mistakes, and repeatedly correcting colleagues consumes not only time but also emotional energy, he said.
“Previously, people who lacked the skills were at least working slowly,” Kim says. “Now we can produce huge quantities in an instant, and seniors need to see it all.”
“Employees are becoming stewards of the machines and of each other’s products,” he added.
The speed of AI has also changed workplace psychology.
Several of the employees interviewed said they felt unable to slow down because AI permanently raised expectations for responsiveness and productivity.
A Seoul-based reporter who requested anonymity said AI tools have changed what editors and managers consider a reasonable amount of work.
“Before AI, people understood that certain tasks took time,” she said. “Now, if something takes more than a few hours, people start asking why.”
Experts warn that companies may be underestimating the long-term human costs of this model.
In an eight-month study, researchers at the Haas School of Business at the University of California, Berkeley, found that employees using AI ended up working faster, longer, and completing more tasks than before.
Harvard Business Review warns that these productivity gains can ultimately lead to lower quality work, higher turnover, and worse decision-making.
“Human cognitive ability is finite,” Hyun says. “AI can operate indefinitely, but humans cannot maintain high-level judgment nonstop.”
For some workers, fatigue is not just about today’s workload, but also the fear of falling behind in a workplace where speed has become the new norm.
Because AI has the potential to widen the productivity gap between high and low performers, some employees fear they will become obsolete faster than ever.
“People feel like they’re always being judged on the speed of AI,” Kim says. “And even those who are surviving now fear that someday they will become one of the slower groups.”
jychoi@heraldcorp.com
