Technostress in the Age of AI: Why Keeping Up Is So Exhausting

“With AI, I did in one hour what used to take me a week.” Messages like this circulate daily on LinkedIn, and leave many readers with an uncomfortable question: “And what am I doing?”

That unease has had a name for forty years: technostress. What artificial intelligence adds is speed: tools that appear every month and rules that change before anyone has learned them.

This article explains what technostress is, why AI intensifies it and what the data say, with the help of a philosopher who saw it coming: Byung-Chul Han, author of The Burnout Society and winner of the 2025 Princess of Asturias Award for Communication and Humanities.

In short

• Technostress is the stress associated with technology. Craig Brod described it in 1984, and Spain’s occupational safety institute, the INSST, treats it as a preventable psychosocial risk.
• According to Byung-Chul Han, today the “you can” tires us more than the “you must”: nobody forces us, we demand it of ourselves. AI multiplies that “you can”.
• AI saves time, but it also consumes it: 77% of those who use it say it has added to their workload (Upwork, 2024).
• The fear of becoming obsolete has grown: from 15% to 22% of US workers between 2021 and 2023 (Gallup).
• Using it more doesn’t always calm things down: in a study of 1,151 people, AI users scored slightly higher on anxiety.
• From fear to fatigue: in the US, searches for “AI fatigue” are now three times those for “AI anxiety” (our own analysis of Ahrefs data).

What technostress is

The term was coined by American psychologist Craig Brod in 1984, in a book with a telling subtitle: The Human Cost of the Computer Revolution. Brod described office workers overwhelmed by the first personal computers.

In Spain, the reference is Technical Prevention Note (NTP) 730 from the National Institute for Safety and Health at Work (INSST), written by Marisa Salanova, Susana Llorens and Eva Cifre, of Universitat Jaume I, together with Clotilde Nogareda. Its approach places the origin outside the person: technostress is a psychosocial harm that can be prevented by the organization.

The five faces of technostress, applied to AI

In 2007, Monideepa Tarafdar and her team published the most cited model on the subject in the Journal of Management Information Systems. They identified five sources of technostress. The model was designed with email and enterprise systems in mind, yet it fits AI almost better than it fit the technology of the time:

Source	What it describes	How it looks with AI
Overload	Technology pushes people to work more and faster	More drafts, more versions, more things to review
Invasion	Work seeps into every moment	The feeling that you could always be automating something
Complexity	The tool is never fully mastered	A new app every month, each with its own logic
Insecurity	Fear of being replaced by technology or by those who master it	”Will AI do this better than me?”
Uncertainty	Constant change leaves no time for learning to settle	What worked in January no longer works in June

What stands out is that none of the five depends on AI working badly: they appear precisely because it works well and changes fast.

Byung-Chul Han: from “you must” to “you can”

In 2010, the philosopher Byung-Chul Han (Seoul, 1959), a professor in Berlin, published a short essay that became a publishing phenomenon: Müdigkeitsgesellschaft, translated into English in 2015 as The Burnout Society (Stanford University Press). Its thesis helps explain why AI is so tiring, even though Han was not writing about it at the time.

From disciplinary society to achievement society

Han starts from a distinction:

• Disciplinary society, the society of the 20th century, ran on prohibitions and external orders: you must, you may not. The boss, the factory, the schedule.
• Achievement society, today’s society, runs on a different verb: you can. Projects, initiative, motivation. Nobody forces you; you force yourself.

The shift looks like liberation, and in part it is. But it comes at a cost: when everything is possible, not doing it is your own fault.

The exploiter who is also the exploited

This is the book’s best-known idea. For Han, the achievement-subject believes itself free, yet demands more of itself than any boss would. Self-exploitation is more efficient than exploitation by others because it comes with a feeling of freedom. Nobody rebels against themselves.

Han links this model to what he calls the neuronal illnesses of the 21st century: depression, ADHD, borderline personality disorder and burnout syndrome. Their origin, he says, lies in an excess of positivity, of possibilities, rather than in an enemy or a prohibition.

AI as the ultimate “you can”

What follows is our own reading, not Han’s: AI takes the “you can” to its extreme. If a tool makes it possible to write, code, design or analyze in minutes, any limit starts to look like an excuse. There is no longer an “I don’t know how” that works as a brake: there is always a tutorial, a prompt or a new app that would solve it.

A boss rarely needs to order anyone to learn the trendy tool. People order it of themselves, at eleven at night, with a “ten AI tricks” video. It is the mechanism Han describes.

Many tabs, little attention

Han devotes another passage to multitasking, and his conclusion is blunt: multitasking, he writes, is no sign of civilizational progress. For him, jumping from one stimulus to another is a broad but shallow form of attention, the kind an animal in the wild needs to keep watch on everything at once. What gets lost is deep attention, the kind that allows us to dwell on something. Anyone who works with three AI assistants open while replying to emails will recognize the scene.

In Non-things (German original 2021; English edition 2022), Han addresses AI directly: he argues that artificial intelligence calculates but does not think, because human thought begins with something that affects and moves us. The thesis is debatable, but it leaves a useful idea: comparing yourself to a machine means measuring yourself with a yardstick that isn’t human.

The paradox: tools that save time and create more work

If AI saves time, why do so many people feel they are working more? History offers precedents.

• In 1930, John Maynard Keynes predicted that his grandchildren would work about fifteen hours a week thanks to technical progress.
• In 1983, historian Ruth Schwartz Cowan showed in More Work for Mother that household appliances did not reduce housework as much as they promised: standards rose (cleaner clothes, washed more often) and tasks that used to be shared ended up concentrated on one person.

With AI, the data point in the same direction. In 2024, the Upwork Research Institute surveyed workers and executives:

• 77% of employees who used AI said it had added to their workload.
• 39% spent more time reviewing what AI generated.
• 23% spent more time learning to use it.
• 21% received more assignments as a direct result of AI.
• 47% did not know how to achieve the productivity their company expected.

Time saved gets reinvested, and rarely in rest.

Outdated if you don’t use it, replaceable if you do

AI also creates a double bind in which neither option is fully reassuring. Microsoft and LinkedIn’s 2024 Work Trend Index, covering 31,000 people in 31 countries, captures it well:

• 52% of those who used AI at work were reluctant to admit using it for their most important tasks.
• 53% worried that using it for those tasks would make them look replaceable.

If you don’t use it, you fear falling behind; if you use it, you fear it will show. Both ways out generate anxiety, which helps explain why so many people use it without saying so.

FOBO: same acronym, different fear

In the article on FOMO we explained that FOBO stands for fear of a better option, the fear that a better plan will come along. At work, the same acronym is used for something else: fear of becoming obsolete.

Gallup measured it in the United States in August 2023:

• 22% of workers feared technology would make their job obsolete, up from 15% in 2021.
• The jump came mainly from college graduates: from 8% to 20% in two years.
• Among workers aged 18 to 34, the figure reached 28%.

For decades automation worried mostly manual work. The jump among college graduates suggests that generative AI has brought that worry to offices, to people who write, analyze or code.

From fear to fatigue: what searches show

For this article we looked at how some Google searches have evolved since 2020, using the monthly estimates from the Ahrefs tool. They are approximations, but they trace a clear trend:

• “AI anxiety” was barely searched in the United States before 2023: between 7 and 15 times a month. After ChatGPT arrived it jumped to 588 in May 2023, and since then it has moved between 400 and 500.
• “AI fatigue” had almost no searches until 2023. It overtook “AI anxiety” for the first time in the summer of 2024, and in 2026 it stands at around 1,200 a month, three times as many.
• “Tecnoestrés” in Spain, the Spanish term for technostress, holds steady at 100 to 180 searches a month. Between 2020 and 2024, August was every year the month with the fewest searches: the term follows the working calendar.

Our reading is that the conversation has moved from fear of what AI might do to the fatigue of living with it. It is the shift Han describes: the exhaustion of someone who can do everything.

How AI anxiety is measured

The most used instrument is the AIAS scale (Artificial Intelligence Anxiety Scale), which Yu-Yin Wang and Yi-Shun Wang published in 2019. It distinguishes four dimensions:

1. Learning: unease about having to learn to use it.
2. Job replacement: the fear of losing one’s job.
3. Sociotechnical blindness: the sense that society is adopting AI faster than it understands the consequences.
4. AI configuration: unease about machines that imitate human traits.

What they found with 1,151 people

In 2025, Naciye Güliz Uğur and Faruk Dursun published a study in Frontiers in Psychiatry with 1,151 adults in Turkey that measured nine dimensions of AI anxiety on a scale from 1 to 5. Some results:

• The highest score went to the desire to learn about AI (4.03). Unease and interest coexist.
• Sociotechnical blindness also scored high (3.41).
• Technophobia, the rejection of technology, was among the lowest (2.86).

Using it more doesn’t always calm things down

It is often said that AI anxiety fades once people use it. The data are less clear:

• In the Turkish study, those who used AI scored slightly higher on overall anxiety than those who did not (3.82 versus 3.60). And those who had been using it for four or five years showed more fear of job replacement than those who had used it for a year or less.
• In a Saudi study with 300 people (Alkhalifah and colleagues, 2024), those who knew about AI but had not tried it showed slightly higher anxiety. The relationship with frequency of use was not significant.

The cautious conclusion is that familiarity is not an automatic antidote. Knowing the tool better also makes it easier to see what it can do.

Technostress and technophobia, two different things

• Technophobia is a rejection of technology. People who have it avoid using it.
• Technostress often appears in people who use technology every day and get a lot out of it. What weighs on them is the pace, the demands and the uncertainty.

This fits the data above: in the Turkish study technophobia scored low, and in Gallup the fear grew mainly among young and college-educated workers. It has more to do with weariness of the familiar than with fear of the unknown.

Who takes it more calmly

Studies do not describe a formula, but they do point to a few traits that recur among people who handle this change better.

Greater tolerance of uncertainty. In 2016, psychologist Nicholas Carleton proposed that fear of the unknown might be the underlying fear behind many others. People who tolerate not knowing how something will turn out suffer less with a technology whose future nobody knows.

Supportive environments. In 2008, Tarafdar’s team identified what reduces technostress in organizations: training on the tool, accessible technical support and worker involvement in how it is rolled out. The environment weighs as much as the person.

A different relationship with achievement. Han closes The Burnout Society with a positive idea he borrows from Peter Handke: alongside the tiredness that isolates, that of someone who exhausts themselves alone chasing their performance, there is a tiredness that unites, one that makes room to stop, look and be with others. For Han, that capacity to stop is exactly what distinguishes us from a machine that calculates.

What we don’t know

• Which jobs will change and when. Predictions about jobs and AI have been revised so often that this article makes none.
• How it is experienced outside the US. Most of the data come from the United States or from global surveys.
• How it evolves over time. Almost all studies are cross-sectional: they measure people at a single moment.
• How far Han goes. He is a philosopher, not an empirical researcher: his essays offer a framework for thinking, not proof.

What does seem solid is the central idea: AI is changing both what we do at work and what we believe we should be doing. For those curious about how they relate to demands and pressure, the burnout test, the perfectionism test and the imposter syndrome test offer an orientation, not a diagnosis. And for anyone whose mind goes blank under stress at the worst moments, the article on why stress blocks memory explains what happens in the brain.

Sources

• Ahrefs, Keywords Explorer. Estimated monthly search volume for “AI anxiety” and “AI fatigue” (United States) and “tecnoestrés” (Spain), January 2020 to October 2026. Our own query of 14 October 2026.
• Alkhalifah, J. M., Bedaiwi, A. M., Shaikh, N., Seddiq, W., & Meo, S. A. (2024). Existential anxiety about artificial intelligence (AI): Is it the end of humanity era or a new chapter in the human revolution: questionnaire-based observational study. Frontiers in Psychiatry, 15, 1368122. https://doi.org/10.3389/fpsyt.2024.1368122
• Brod, C. (1984). Technostress: The Human Cost of the Computer Revolution. Reading (Massachusetts): Addison-Wesley.
• Carleton, R. N. (2016). Fear of the unknown: One fear to rule them all? Journal of Anxiety Disorders, 41, 5-21. https://doi.org/10.1016/j.janxdis.2016.03.011
• Cowan, R. S. (1983). More Work for Mother: The Ironies of Household Technology from the Open Hearth to the Microwave. New York: Basic Books.
• Gallup (2023, September). More U.S. workers fear technology making their jobs obsolete. https://news.gallup.com/poll/510551/workers-fear-technology-making-jobs-obsolete.aspx
• Han, B.-C. (2015). The Burnout Society (E. Butler, Trans.). Stanford: Stanford University Press. (German original: Müdigkeitsgesellschaft, 2010).
• Han, B.-C. (2022). Non-things: Upheaval in the Lifeworld (D. Steuer, Trans.). Cambridge: Polity. (German original: Undinge: Umbrüche der Lebenswelt, 2021).
• Keynes, J. M. (1930). Economic possibilities for our grandchildren. In Essays in Persuasion (1931). London: Macmillan.
• Microsoft & LinkedIn (2024). 2024 Work Trend Index Annual Report: AI at Work Is Here. Now Comes the Hard Part.
• Ragu-Nathan, T. S., Tarafdar, M., Ragu-Nathan, B. S., & Tu, Q. (2008). The consequences of technostress for end users in organizations: Conceptual development and empirical validation. Information Systems Research, 19(4), 417-433.
• Salanova, M., Llorens, S., Cifre, E., & Nogareda, C. (2006). NTP 730: Tecnoestrés: concepto, medida e intervención psicosocial. Madrid: Instituto Nacional de Seguridad e Higiene en el Trabajo. https://www.insst.es/documentacion/colecciones-tecnicas/ntp-notas-tecnicas-de-prevencion/21-serie-ntp-numeros-716-a-750-ano-2006/ntp-730-tecnoestres-concepto-medida-e-intervencion-psicosocial
• Tarafdar, M., Tu, Q., Ragu-Nathan, B. S., & Ragu-Nathan, T. S. (2007). The impact of technostress on role stress and productivity. Journal of Management Information Systems, 24(1), 301-328.
• Upwork Research Institute (2024, July). Upwork study finds employee workloads rising despite increased C-suite investment in artificial intelligence. https://www.upwork.com/press/releases/upwork-study-finds-employee-workloads-rising-despite-increased-c-suite-investment-in-artificial-intelligence
• Uğur, N. G., & Dursun, F. (2025). The social anatomy of AI anxiety: gender, generations, and technological exposure. Frontiers in Psychiatry, 16, 1641546. https://doi.org/10.3389/fpsyt.2025.1641546
• Wang, Y.-Y., & Wang, Y.-S. (2022). Development and validation of an artificial intelligence anxiety scale: an initial application in predicting motivated learning behavior. Interactive Learning Environments, 30(4), 619-634. Published online in 2019. https://doi.org/10.1080/10494820.2019.1674887