A Debate about the Displacement Shadow: Fear of AI Replacement and Workplace Conduct

[00:00:00] Welcome to The Debate. Right now, in offices around the world, employees are showing up late. They're scrolling through their phones for hours on the clock. They are ignoring deadlines and they're misusing company resources. And they are doing all of this because of a machine that hasn't even taken their job yet. We are essentially watching people react to a ghost. Exactly. I mean, think about how a meteorologist forecast a catastrophic storm.

[00:00:27] The destruction on the ground actually doesn't start when the hurricane makes landfall. It begins days, sometimes weeks before. Right. People panic. Yeah. They hoard resources. They abandon their posts. They stop performing routine maintenance because they figure, well, what is the point of fixing the roof if the whole house is going to blow away? The mere anticipation of the event causes real, measurable damage long before the first drop of rain falls. So the psychological forecast creates a physical reality in the present. We act on the future we anticipate.

[00:00:56] And we are seeing that exact same meteorological phenomenon in the workplace right now with artificial intelligence. Researchers call it the displacement shadow. There's this fascinating study from October 2026 by Probst and colleagues. Oh, right. The one across the 18 industries. Yes, exactly. They surveyed workers across 18 different industries and found that the mere anticipation of AI-driven job obsolescence is actively eroding everyday workplace conduct. When employees fear their work is disappearing, they engage in moral disengagement.

[00:01:25] They basically grant themselves silent permission to cut corners. Right. So the central question for us today is this. When employees fear AI will make their skills obsolete and they rationalize these counterproductive behaviors, is this behavioral erosion fundamentally a failure of leadership to manage the transition? Or is it an inevitable structural psychological response to true occupational insecurity that organizational policies can only, you know, marginally contain?

[00:01:49] While the fundamental driver here is existential, I will argue that the root cause is structural exposure to AI, meaning that this moral disengagement is a cognitive coping mechanism that happens regardless of corporate messaging. And I am going to argue that this erosion of conduct is primarily a leadership failure born from misbanded communication, broken psychological contracts, and just a lack of credible redeployment pathways. Okay, then let's establish exactly what is happening in the minds of these workers.

[00:02:16] We have to understand that what we are dealing with right now is not traditional job insecurity. Right. It is occupation insecurity. Job insecurity is the fear of losing your position at your current company. You can fix that by updating your resume, maybe taking a few interviews, finding a new employer. Sure, yeah. Occupation insecurity, which researchers like Roll and colleagues validated a few years ago, is the profound existential fear that your entire line of work is disappearing permanently. But that fear doesn't exist in a vacuum.

[00:02:45] I mean, it is heavily influenced by how leaders frame the future. The harms we are seeing from AI anticipation, this displacement shadow, stem directly from organizational mismanagement. I disagree. Well, look, we have decades of research showing that employees do not just react negatively to bad outcomes. They react to perceived unfairness. They react to broken promises. There's this concept called the psychological contract. Right, the unwritten expectations. Exactly.

[00:03:09] It's the silent agreement between an employer and an employee that basically says, if I give you my loyalty and hard work, you will give me stability and a path forward. Right now, leaders are actively breaching that contract. But hearing reassuring words from a CEO does not fix an expiring profession. Think about the heavy AI users. The 2026 Workday survey found a massive trend. Heavy AI users actually have the highest anxiety. Which seems counterintuitive at first. Exactly. Why does using the tool make them more anxious?

[00:03:39] Because every time they ask the AI to draft a complex financial report in 10 seconds, they realize, um, I went to school for four years to learn how to do this. The tool is a daily mirror reflecting their own obsolescence. That is a structural problem. It completely transcends individual leadership tactics. I think you are severely underestimating how much leadership tactics fuel that exact anxiety, though. The displacement shadow is cast by leaders who prioritize external investor narratives over internal transparency.

[00:04:09] Let's look at how this actually takes root on the ground. Okay. Let's talk about the information vacuum. Look at IBM back in 2023. The CEO publicly commented that roughly 30% of their back office roles could be replaced by AI over five years. Yeah, he said that right out loud. Right out loud. And predictably, this created massive internal panic. Corporate spokespeople had to furiously walk back the comments, clarifying there was no blanket hiring pause. But the damage was done at that point.

[00:04:37] Because treating your AI strategy as an investor relations tool to, you know, boost your stock price, while completely ignoring how your employee audience will hear it, is like casually discussing a home demolition in front of the people currently living in the house. That's a dramatic way to put it. But it's true. Contrast that with Walmart a bit later. Their CEO explicitly acknowledged that AI would change every job, but paired that reality with a clear message.

[00:05:02] He said global headcount would remain flat, and they were immediately launching AI certification programs. The anxiety isn't structural. It's proportional to how recklessly the message is delivered. Okay, but I just don't buy that it's merely a messaging problem. You mentioned IBM and Walmart as if this is just a PR puzzle to be solved. Employees are not reacting to botched PR. They are reading objective macroeconomic signals. Which are often inflated by the companies themselves. Take a look at the aggregate data, though.

[00:05:31] In just the first seven months of 2026, we saw over 100,000 announced job cuts directly attributed to AI. Hold on. We have to contextualize those numbers. How many of those 100,000 cuts were purely driven by actual AI capability? And how many were just macroeconomic belt tightening? Where companies used AI efficiency as a convenient excuse for Wall Street? Companies love to blame the algorithm when they just want to slash payroll.

[00:05:55] Even if some companies are hiding regular layoffs behind the AI buzzword, the aggregate signal to the worker is deafening. You cannot PR your way out of the fact that the market is shrinking for human labor in these sectors. It is not an information vacuum. It is an information surplus. Let me give you an example. When the CEO of Duolingo posted a memo about stopping the use of human contractors because AI could handle the translation work, he faced public backlash and later claimed he didn't provide enough context. But the employees understood the context perfectly well.

[00:06:24] The machines can do the work now. The threat isn't a communication error. It's real. Yes, the market reality is shifting. I'll give you that. But it's how organizations mediate that reality that dictates whether an employee steals time or stays engaged. If we look back at the probe study on moral disengagement, we really need to understand the mechanism here. Go on. Moral disengagement means an employee feels they are being treated unjustly, so they loosen their own moral self-regulation to compensate.

[00:06:51] They think, well, the company is screwing me, so taking two-hour lunches is fair game. The antidote to that feeling of injustice is a credible redeployment commitment. Look at what IKEA did with their call centers. Right. They rolled out an AI assistant that handled routine customer inquiries. Exactly. Routine work that used to be done by thousands of people. But IKEA didn't just fire those 8,500 call center workers. Instead, they explicitly chose to rescale those exact workers as remote interior design advisors. Which is great. It is.

[00:07:21] They created a massive new revenue stream while preserving the workforce. Or, you know, look at JPMorgan Chase and highly exposed finance. Their leadership made an explicit public commitment to retrain and redeploy affected employees. If companies build tangible pathways, they neutralize the psychological contract breach. Why wouldn't explicit funded redeployment commitments prevent the behavioral erosion we are talking about? Because reskilling has a cognitive ceiling. And we are completely ignoring it.

[00:07:50] I will absolutely acknowledge IKEA's success. It is a fantastic outcome. Thank you. But it is also a best-case scenario where routine query work translates neatly into human-centric advisory work. The skills are adjacent. Now, look at Accenture's restructuring in 2025. They were rapidly growing their AI headcount, but their CEO explicitly admitted that 11,000 staff members had to be exited. Why? Because their specific skills simply could not be updated through reskilling to match the new AI paradigms.

[00:08:18] Wait, so you're saying they just couldn't be taught the new systems? I'm saying that occupation insecurity is like being on a sinking ship. Offering a reskilling program that only has room or only has the technical capacity for a fraction of the passengers doesn't stop the panic. It just changes who gets a lifeboat. That's pretty bleak. It's realistic. We are talking about mid-career and late-career professionals. Imagine being a 45-year-old actuary.

[00:08:42] The psychological burden of realizing your hard-earned expertise is structurally obsolete cannot be patched over by HR sending you a link to an upskilling portal. You cannot reskill your way out of a profound identity crisis. But that assumes that if we cannot save every single job, we have no mechanism to maintain trust. And that simply isn't true. Let me introduce a concept here. Procedural justice. Okay. It is essentially a fancy way of saying, even if I don't like the outcome, I feel like the process was fair.

[00:09:12] The historical research on procedural justice proves that process matters just as much as outcomes. If we go back to classic field studies in the 1990s, researchers found that when employee pay was cut, employee theft skyrocketed. Naturally. But that only happened in the group where managers failed to explain the reasons respectfully. In the group where the process was transparent and respectful, misconduct barely rose. So you're arguing that how you fire someone dictates how the survivors behave? 100%. Fair process preserves the ethical climate.

[00:09:43] Take Dropbox a few years ago. They cut 16% of their workforce, explicitly citing the shift to the AI era. But they provided 16 weeks of severance, advanced notice, and deep transition support. They managed the fairness of the exit. Furthermore, for the survivors, organizations must provide voice. Voice? Yeah. Look at the Writers Guild of America strike in 2023. By fighting for and co-designing guardrails against AI-generated source material, writers regained a sense of control.

[00:10:12] Giving employees a voice and fostering strong, honest manager relationships prevents the exact feeling of powerlessness that triggers counterproductive work behavior. Okay. This is where we need to stop. Because this is where the data gets incredibly counterintuitive. Voice and procedural justice are wonderful concepts. But when we look at the empirical data on how managerial empathy operates in the shadow of AI, we run into a very dangerous trap. It is called the empathy paradox. The empathy paradox? Yes.

[00:10:40] You are arguing for strong manager relationships and support. But let's look at the 2020 Prope Study on Safety-Related Moral Disengagement. That study found that perceived organizational and supervisor support actually made the problem worse. Worse? Yes. If a manager was highly supportive, it strengthened the indirect effect of job insecurity on poor performance, unless it was paired with incredibly strict standards. Wait, wait. Being a supportive manager caused more rule-breaking? How does that work? Think about the mechanism.

[00:11:08] If you are a highly empathetic manager, you sit down with your employee. You validate their fears about AI. You tell them, look, I know this transition is rough. I know you're worried about your job. But if you don't simultaneously enforce strict accountability, the employee reads that empathy as permission. Oh, I see. They think, well, my manager knows I'm getting a raw deal with this AI rollout. They know I'm stressed, so they won't mind if I browse the web for two hours today. It exacerbates the very misconduct you are trying to solve.

[00:11:36] Empathy without enforcement creates a license to disengage. I mean, that relies on managers completely dropping the ball on basic management. It relies on managers being human. You praise CEOs for making public pledges to protect their workers. But what happens to an ethical climate when a CEO promises no AI layoffs, like Salesforce essentially did in early 2025, only to cut thousands of support jobs months later because AI agents reduced the need for humans? But that... The inconsistency shatters whatever relational trust existed.

[00:12:06] The structural reality of AI efficiency will always eventually overpower empathetic pledges. But the Salesforce example perfectly proves my point. The behavioral damage inside Salesforce wasn't caused by the AI agents. It was caused by the CEO's hypocrisy. It was a failure of leadership to maintain a credible narrative. When leaders say one thing and do another, they provide the exact rationalization employees need to cognitively detach. I hear you, but... The core of my argument remains.

[00:12:33] While AI displacement is undeniably a massive shift in how work is done, the organizational response is what determines the behavioral fallout. Look, if an entire department's function can be automated, no amount of leadership consistency will stop the employees in that department from checking out mentally before the axe falls. I strongly disagree. Because the data clearly shows a massive gap in conduct between organizations with high procedural justice and low procedural justice.

[00:12:58] If you are a manager listening to this, you're probably thinking, but what can I actually do if I don't control the technology? Right. You can create transparent roadmaps. You can be honest about uncertainty rather than making impossible promises you'll have to break in six months. You attempt redeployment before layoffs, and you structurally share the productivity gains of AI with the remaining workforce. The displacement shadow is dark, but the organization controls the light. I agree that fairness is important, and I certainly agree that leaders shouldn't make contradictory promises.

[00:13:27] But leaders must also accept the limits of their influence here. The displacement shadow is cast by the technology, not the manager. We are talking about occupation insecurity. It's a profound identity threat that fundamentally destabilizes professionals. Because AI exposure, the literal percentage of overlap between human tasks and what a large language model can do, so heavily dictates the level of moral disengagement, organizations must realize they are managing a structural psychological crisis.

[00:13:55] So you really believe it's just beyond the scope of HR? Entirely. The recent data shows that the highest moral disengagement is happening in the most heavily exposed industries. It transcends individual corporate cultures. When an employee knows the machine can do their job better, faster, and cheaper, the psychological defense mechanism kicks in. It is not a procedural HR issue to be smoothed over with a nice severance package and an upskilling module. It is a fundamental rewiring of the human relationship to labor.

[00:14:22] Well, while we clearly disagree on the root cause, whether this behavioral erosion is a crisis of managerial process or an unavoidable symptom of technological revolution, I think we both firmly agree on the immense stakes involved. Oh, definitely. We will see that AI anticipation has real measurable business costs that extend far beyond a few people taking long lunches. Absolutely. We are seeing it impact physical safety compliance. We are seeing massive retention issues with top talent who refuse to wait around to be replaced.

[00:14:52] And we are seeing a direct hit to service quality. Exactly. A perfect example of that service quality impact is Klarna. They rolled out an AI assistant that handled two-thirds of their customer service chats. They bragged that it was doing the equivalent work of 700 full-time human staff. Right. But by 2025, their leadership had to acknowledge that focusing purely on cost automation caused the quality of complex issue resolution to suffer so severely that they actually had to start recruiting human agents again. Customers just wanted to reach a person.

[00:15:22] The micromanagerial responses, how you deploy the tech and how you value the humans interacting with it, matter just as much as the macro technological reality. It is a complex puzzle, and it demands that we look at both the human psychology of an expiring identity and the cold, hard data of technological exposure without flinching. Well said. We'll leave it to you, the listener, to decide where the ultimate responsibility lies. But remember our forecast. You don't have to wait for the hurricane to make landfall to see the damage. The shadow of the storm is already here.

[00:15:50] It is changing how we work, how we trust, and how we behave every single day. Thank you for joining us on The Debate.