This is the second part of a two-part series on the AI Red Queen race. Part I examined why the corporate AI race produces no lasting winners. Part II asks what happens to the people caught inside it.
There is a question that most AI discourse carefully avoids. Not whether AI will take jobs, or how many, or how fast. Those are forecasting questions, and forecasting is comfortable because it keeps the problem in the future.
The harder question is already here, in the present tense: if a machine can do what I do, then what is my value?
This is not an abstract philosophical puzzle. It is the question that an accounts payable specialist asks herself when the new AI tool processes invoices she used to spend a full day on. It is the question a middle manager faces when a dashboard now routes the information he used to carry between teams. It is the question a junior analyst confronts when a senior colleague, armed with an AI assistant, produces in an afternoon what used to take the entire team a week.
The question does not arrive gently. And it is about to arrive for a great many people at once.
The workforce we are starting with
Before examining what AI does to the people inside organizations, it is worth pausing on the starting condition. Because the workforce that is about to be reshaped by AI is not, by any measure, in good shape.
Gallup’s 2026 State of the Global Workplace report, based on 2025 data, found that only 20% of employees worldwide are engaged at work. Sixty-four percent are not engaged, meaning they show up, do the contractual minimum, and are psychologically detached. Sixteen percent are actively disengaged, meaning they are not just checked out but working against the organization’s interests.¹
In Europe, the picture is worse. Only 13% of workers are genuinely engaged. The continent that prides itself on worker protections and social dialogue has, by this measure, the least invested workforce in the developed world.²
These numbers predate the AI restructuring wave. This is the baseline. The workforce that companies are now asking to “transform” and “adapt” is a workforce where roughly two out of three people have already retreated into minimum compliance.
The phrase that entered the popular vocabulary for this is “quiet quitting.” It sounds like laziness. The research says otherwise.
A rational retreat
Organizational psychologists have a more precise term for what is happening: psychological contract breach. The concept is simple. Every employment relationship rests on an implicit deal, rarely written down but deeply felt. The employee gives loyal effort, initiative, flexibility. In return, the employer provides stability, respect, growth, a sense that the contribution matters. When employees perceive that this deal has been broken, through layoffs, broken promises, shifting expectations, or management perceived as toxic, they reduce their investment. Not out of spite. Out of rational self-protection.
The data on this is not ambiguous. A meta-analysis on psychological contract breach found that perceived violations correlate with a 37% drop in discretionary effort and a 28% decline in organizational citizenship behaviors, the kind of voluntary cooperation, helping colleagues, going beyond the strict job description, that keeps organizations running.³
What the popular discourse calls “quiet quitting” is, in research terms, a calculated withdrawal of discretionary effort in response to a perceived broken contract. It is not a moral failing. It is a structural outcome.
And it was already the dominant mode of work for the majority of the global workforce before AI entered the picture.
The moment it stops being optional
In Part I, I described the theater of AI readiness: the small task force that experiments while 95% of the organization continues unchanged. That phase is comfortable for everyone. The enthusiasts get to explore. The rest of the company gets to ignore it. Nobody’s value is questioned.
That phase is ending.
At Block, Jack Dorsey did not stop at cutting 40% of the workforce. He mandated that every remaining employee use AI tools daily. AI fluency was formally integrated into performance evaluations. Employees are now required to send Dorsey a weekly email listing their five most recent accomplishments, which he processes using AI summaries.⁴ The message is not subtle: use AI or become the next position to be eliminated.
This is the moment where AI stops being an interesting experiment and becomes an existential pressure applied to every person in the building. It is no longer “here is a tool, try it if it helps.” It is “this is how we work now, and the way things are measured has changed.”
For the person who spent fifteen years building professional identity around mastering a particular software, a specific reporting process, or a domain of administrative expertise, this is not a workflow adjustment. It is a statement that the skill that defined their professional worth is no longer sufficient. The question beneath every “AI transformation” initiative is one that no change management framework says out loud: we are about to find out which parts of this role were the person’s unique contribution and which parts were just process.
For a significant number of people, the honest answer will be uncomfortable.
What happens to those who stay
The research on what follows is extensive, consistent, and largely ignored by the people making restructuring decisions.
Workplace survivor syndrome, the documented psychological response of employees who remain after layoffs, produces effects that are the opposite of what the restructuring was meant to achieve. A study of over 4,000 employees who survived layoff rounds found that 74% reported their own productivity had declined, 69% said the quality of the product or service had dropped, and 81% judged that customer service had deteriorated.⁵ Other research estimates a roughly 20% performance decline among survivors when nothing is done to rebuild trust.⁶
The mechanism is exactly what the psychological contract research predicts. Remaining employees watch colleagues get cut, observe that loyalty and effort did not protect anyone, and draw the rational conclusion: invest less. Do what is contractually required. Volunteer nothing. Take no risks. Protect the perimeter.
The company runs faster on paper. The humans inside it slow down. The net movement, in Red Queen terms, is zero. Possibly negative.
The efficiency gains were, in most documented cases, not what the spreadsheet projected.
The individual Red Queen
But the damage extends beyond survivor syndrome. There is a subtler, slower dynamic at work, one that applies not just to companies that have gone through layoffs but to every workplace where AI is raising the bar.
When AI makes it trivially easy to produce volume, volume becomes worthless as a signal. The developer who could write fifty lines of clean code per day was once clearly valuable. When an AI-assisted colleague ships two hundred lines, the old standard evaporates. The analyst who built a financial model in three days watches a junior colleague produce something comparable in three hours. The content writer who was the team’s most productive voice finds that two people with AI tools now match what five used to produce.
The output expected rises. The compensation does not. The definition of “enough” keeps moving. This is the Red Queen operating at the individual level, and it is perhaps the cruelest version of the race, because there is no team, no strategy, no competitive positioning to hide behind. It is just a person, running faster, watching the bar rise, wondering how long they can keep up.
And beneath the running lies that quiet, corrosive question: if the machine produces the same output, what was I contributing in the first place?
The load-bearing wall
Here is where the argument moves from organizational to structural, and where the stakes become genuinely serious.
In OECD countries, roughly 72% of the working-age population is employed.⁷ The entire architecture of modern economies, the consumer spending that drives GDP, the tax base that funds public services, the pension systems, the housing markets, the healthcare models (especially in the United States, where insurance is tied to employment), all of it rests on one foundational assumption: most adults work, earn, spend, and pay taxes through stable employment.
This is not an ideological point. It is a load-bearing wall.
Gallup estimates that disengagement already costs the global economy over $10 trillion per year in lost productivity, representing 9% of global GDP.⁸ That is the cost of a workforce where two-thirds of people are psychologically checked out. Now consider what happens when AI simultaneously raises the performance bar (making every individual’s output feel less special), triggers restructuring waves (activating survivor syndrome across entire industries), and forces a confrontation with the value question (making millions of people wonder whether their role is fundamentally process that a machine can replicate).
The disengagement does not stay inside the office. Gallup’s own data shows the link between work engagement and overall life satisfaction: half of engaged employees describe themselves as thriving, compared to only a third of the disengaged.⁹ When the workplace ceases to be a source of meaning, identity, and social connection, and becomes instead a treadmill where the pace always increases and the purpose always recedes, the effects ripple outward into families, communities, consumer behavior, political sentiment, and social cohesion.
The system we have was designed for a world where most people have stable, moderately meaningful work. If AI turns that work into a permanent Red Queen race, always running, never arriving, the system does not adapt gracefully. It was not built to.
What the Queen never explains
In Through the Looking-Glass, the entire story takes place on a giant chessboard. The pieces move according to rules they do not fully understand, toward outcomes they did not choose. Alice eventually reaches the eighth square and becomes a Queen herself. But the moment she is crowned, the dream ends. She wakes up. Carroll never tells us whether waking up was the victory or the loss.
We have spent a great deal of energy debating whether AI will take jobs. The forecasts vary. The timelines shift. The consulting decks get updated quarterly.
The more uncomfortable question, the one that sits beneath all the projections and restructuring plans and transformation roadmaps, is not about jobs. It is about what those jobs meant. To the people who held them. To the families that depended on them. To the societies that were built, brick by brick, on the assumption that work gives people not just income but purpose, structure, and a reason to believe that their effort matters.
If the Red Queen’s race hollows that out, if it turns work into an endless acceleration with no destination and no reward beyond survival, then the question is not whether the corporate world adapts.
The question is what it adapts into.
References
- Gallup, “State of the Global Workplace: 2026 Report” (data collected January-December 2025), published April 2026. https://www.gallup.com/workplace/697904/state-of-the-global-workplace-global-data.aspx
- Gallup, “State of the Global Workplace: 2025 Report” (data collected 2024); Europe engagement at 13%. Confirmed in the 2026 report. https://healthyworkcompany.com/wp-content/uploads/2025/05/state-of-the-global-workplace-2025-download.pdf
- Meta-analysis on psychological contract breach, cited in: Sfeir, Ph.D., “The Hidden Cost of Psychological Contract Breach,” LinkedIn, 2025. Original research: Zhao, H., Wayne, S.J., Glibkowski, B.C. & Bravo, J. (2007), “The Impact of Psychological Contract Breach on Work-Related Outcomes: A Meta-Analysis,” Personnel Psychology, 60(3), 647-680.
- Metaintro, “Jack Dorsey Faces Employee Backlash After Block Layoffs and AI Mandates,” February 27, 2026. https://www.metaintro.com/blog/jack-dorsey-block-layoffs-ai-mandates-employee-backlash-2026
- LeadershipIQ, “Don’t Expect Layoff Survivors to Be Grateful.” https://www.leadershipiq.com/blogs/leadershipiq/29062401-dont-expect-layoff-survivors-to-be-grateful
- Multiple sources converge on the ~20% figure: Stockholm School of Economics, “Dealing with Survivor Syndrome,” 2024; ImpactGroup HR, “Survivor Syndrome,” 2024. https://www.hhs.se/en/research/sweden-through-the-crisis/dealing-with-survivor-syndrome/
- OECD, Employment Outlook 2025: average employment rate rose to 72.1% in Q1 2025. https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/07/oecd-employment-outlook-2025_5345f034/194a947b-en.pdf
- Gallup, “Global Employee Engagement Falls to Lowest Level Since 2020,” press release, April 8, 2026. “In 2024 alone, disengagement resulted in more than $10 trillion in lost productivity globally, representing 9% of global GDP.” https://www.prnewswire.com/news-releases/global-employee-engagement-falls-to-lowest-level-since-2020-302732859.html
- Gallup, “State of the Global Workplace: 2025 Report.” Thriving data: 50% of engaged employees describe themselves as thriving vs. 33% of non-engaged.
