The AI Anxiety Era: What AI Worry at Work Is Costing American Workers

American workers are absorbing the mental health cost of AI on their own, and more are turning to a chatbot for support than to a clinician.

One in 4 U.S. layoffs this year has been blamed on AI, making it the leading reason employers gave for cutting jobs. Workers see those headlines the same day their managers do. Then, they return to a desk where their managers think that using AI to draft an email means their workers have an extra 10 hours in the week – and assign tasks accordingly. This situation becomes untenable for many, causing them to worry about layoffs, being replaced by an AI agent, and whether or not they’ll ever find a job that won’t rely heavily on AI.

To understand what that pressure is doing to people, The Sylvia Brafman Mental Health Center Georgia surveyed 1,000 U.S. adults about their experience of artificial intelligence at work: how secure they feel, how their workload has shifted, what they are doing after hours to stay employable, and where they turn when the worry follows them home. The answers describe a workforce absorbing a great deal quietly and on its own. Younger workers carry the heaviest share of it – and more people are taking their AI anxiety to a chatbot than to a professional.

Key Takeaways

• Half of workers (49%) said worrying about AI at work has taken a toll on their mental health or well-being—but only 13% of them have talked to a therapist, counselor, doctor, or other professional about it.
• 9% turned to an AI chatbot for emotional support or reassurance about AI at work—more than the 6% who went to a human mental-health professional. Among Gen Z, chatbot use climbs to 17%.
• 56% are concerned AI could replace some or all of their work within five years—rising to 66% of Gen Z and falling to 38% of Boomers.
• 64% have done something in the past year to prove their value or avoid looking replaceable.
• 71% of Gen Z say AI worry has affected their mental health or well-being, compared to 28% of Boomers, and 78% of Gen Z have done at least one thing to make their job more secure, versus 46% of Boomers.
• Nearly 1 in 5 (19%) avoided asking for help at work so they wouldn’t seem replaceable—rising to 28% among Gen Z. Another 17% worked while sick for the same reason.
• 42% say AI cuts the time their tasks take —but 33% said this led to a higher workload once AI was implemented at work.
• 51% applied for or seriously considered a new job for an AI-related reason in the past year—jumping to 77% of Gen Z and dropping to 27% of Boomers.

Workers Are Handling AI Anxiety on Their Own

Layoff announcements rarely use the word AI. They say restructuring or efficiency, and the people reading them fill in the rest. That worry follows a worker – right into the meeting where their boss wonders whether the deliverable they spent two days on could have been produced in a minute.

Across the effects measured, 49% of workers said AI worry had affected their mental health or well-being in at least one way. The most common disruptions included:

• Increased stress or anxiety (17%)
• Less confident in professional abilities (16%)
• Less motivated at work (15%)
• Emotionally exhausted or burned out (14%)
• Worried about paying bills or maintaining financial stability (13%)
• Trouble sleeping (12%)
• Trouble concentrating (11%)
• Irritability or mood changes (11%)
• Trouble separating work from personal life (10%)

Not all generations experienced this equally. While 71% of Gen Z reported at least one of the above disruptions, only 46% of millennials, 35% of Gen X, and 28% of baby boomers felt the same. Younger workers doubted their abilities much more than their older counterparts: 28% of Gen Z felt less confident in their professional skills because of AI, compared to only 6% of boomers. Those who entered the workforce most recently are watching entry-level positions turn into AI agents, and it’s making them doubt the future of their entire career.

More workers took their AI anxiety to an AI than to a person trained to treat it. A minority of workers (9%) turned to a chatbot for emotional support or reassurance, but even fewer (6%) went to a therapist, counselor, doctor, or other professional. Among workers who reported AI-related stress, 77% went to neither a professional nor a chatbot.  Gen Z was most likely to seek help, with 17% utilizing chatbots and 12% turning to professionals, the largest share of any generation. With mental health support becoming more normalized for younger Americans, Gen Z is more comfortable seeking support for their stress and anxiety.

Men turned to a chatbot at 12% against 5% of women, while 6% of each gender saw a professional. Women were more concerned with their finances, with 16% worried about paying bills against 11% of men. In an era where more women than ever are the financial head of their households, they shoulder more of the burden if AI makes their jobs redundant.

Younger Workers Feel Most Replaceable

Workers at the end of their careers can use their decades of experience to prove AI can’t replace them. Younger Americans, who have only been in the workforce for a few years, don’t have the same fallback.

Over half (56%) of respondents were concerned that AI could replicate their work within the next five years, including 15% who were extremely concerned. Gen Z was once again the most worried: 66% of Zoomers were worried about being replaced by AI, followed by 57% of millennials, 51% of Gen X, and 38% of baby boomers. Younger workers are feeling this worry more acutely, with a quarter (24%) of Gen Z saying they are extremely concerned about being replaced by AI. Workers at the start of their careers have the least accumulated proof of their own value, and the tasks that build that proof are the first ones to be automated.

Because of these worries, many workers feel destabilized. A third of workers (30%) felt less secure about their job or career than they did a year ago, including 6% who felt much less secure. Only 20% said they felt more secure in their careers than a year ago. The generational divide here runs deeper: 42% of Gen Z feel less secure than a year ago, double the 21% of boomers who agreed. Meanwhile, 51% of boomers reported no change, compared to 27% of Gen Z. While the older generation feels that nothing is changing their careers, Gen Z is grappling with the fact that their career paths may never be the same.

Women carried more of this worry than men. A third of women (33%) were extremely or very concerned that AI could replace their work, compared to 26% of men, and 34% felt less secure than a year ago, compared to 26% of men. Women experience impostor syndrome more frequently than men, meaning that even if they are essential to their teams at work, they may struggle to believe it.

Men were also more likely to say the past year had improved their position, with a quarter (25%) feeling more secure compared to 16% of women. Women have been historically vulnerable to layoffs and job cuts, and that shapes how they feel about technology built to do their work.

AI Saves Time, But Not For Workers

Widespread AI adoption happened faster than many workers were prepared for, but many admitted it can save time. While executives may overstate the time-saving benefits because they’re impressed with the technology, workers report that AI can be a powerful tool in their arsenal.

Over two in five workers (42%) said AI reduces the time their tasks take, including 11% who described the reduction as substantial. Another 18% said AI has little or no effect on how long their work takes, 14% said it actually adds time, and 26% do not use AI for work-related tasks at all. The gains are real but modest for most people who see them.

Expectations moved faster than the workers could. A third of workers (33%) said expectations for how much they should complete went up once AI was adopted, against 7% who said expectations went down. Even if the tools do free up an employee’s time, that time is immediately filled by new tasks and responsibilities.

Workers are watching that trade play out across their organizations. Here are the most common changes workers reported after AI came into their workplace:

• Employees expected to complete more work using AI (27%);
• Responsibilities reduced or reassigned after AI was introduced (24%);
• Coworkers relying on AI-generated work without checking it for accuracy (18%);
• Employees told to use AI without adequate guidance or training (17%);
• Layoffs or role eliminations linked to AI or automation (16%);
• Roles left unfilled because technology could absorb the work (15%);
• People using AI excessively to appear more productive (15%).

A third of workers (33%) have seen none of these changes, and 24% said AI tools have not been introduced or widely used at their workplace at all.

Income determines who absorbs the pressure. A third of workers earning $150,000 or more (34%) have been expected to complete more work using AI, compared with 20% of those earning under $50,000. But while the higher earners are seeing these changes, they’re not the ones making up the difference: 46% of the highest earners said expectations for their own output did not change at all, more than double the amount of workers earning under $50,000 who said the same (22%).

Workers Are Paying to Keep the Jobs They Have

To keep up, American workers have to stay late and hustle harder. Many are seeking additional training off the clock just to stay competitive in the job they were already hired for.

Over half of workers (51%) have felt pressure to improve their AI skills to stay competitive, and 9% feel it very often. Another 30% have never felt it at all. However, who those workers are matters more than the topline: the “never” figure runs from 12% among Gen Z to 38% among boomers, meaning the pressure is concentrated almost entirely on younger workers.

That pressure makes workers put in time off the clock. More than a quarter of workers (27%) have practiced using AI tools outside normal work hours, and 24% completed an AI course, certification, or training in the past year. Another 12% paid personally for AI software, training, or education, spending their own money to remain qualified for a job they already have. Nearly a fifth (18%) changed how they describe their skills on a resume or professional profile because of AI, and 21% have considered changing roles, careers, or industries entirely. Another 9% delayed a career move because they felt unprepared for AI-related expectations.

Younger generations were also making more moves to stay competitive. A third (32%) of Gen Z completed AI courses, against 12% of boomers. Millennials were the most likely to pay for their training out of pocket, at 19% compared to 12% of Gen Z and 6% of boomers. Men outpaced women on both effort and anxiety: 30% of men completed a course or certification against 18% of women, and 46% of men worry that coworkers using AI more have an advantage over them, against 36% of women. Nobody has agreed on what AI competence looks like yet, which makes a certification easy to obtain and hard to quantify. The gap may say more about who feels compelled to document their skills than about who has them.

What Workers Do to Seem Indispensable

When company leadership begins seeing positions as something an AI agent can do, workers notice. In response, American workers find ways to make sure their bosses know a computer could not do their job alone. But doing so adds stress to workers’ lives.

Nearly two-thirds of workers (64%) have done something in the past year to prove their value or avoid looking replaceable. No single behavior accounts for that number. It comes from a cluster of small ones stacked on top of each other:

• Made themselves more available to managers or coworkers (22%);
• Worked beyond scheduled hours specifically to avoid appearing replaceable (21%);
• Answered work messages outside normal hours (20%);
• Took on additional tasks without additional pay (19%);
• Avoided asking for help so they would not seem replaceable (19%);
• Worked while sick for the same reason (17%);
• Delayed or skipped taking time off (16%);
• Avoided setting any limits on their workload (15%);
• Overstated their AI skills or claimed AI use they had not done (15%).

These behaviors become persistent. Among workers reporting at least one behavior, 35% do it sometimes, 20% often, and 14% very often – meaning a third of that group has made it a habit.

Over three-quarters (78%) of Gen Z reported at least one of these behaviors, followed by 64% of millennials, 54% of Gen X, and 46% of baby boomers. Over a quarter of Gen Z (26%) overstated their AI skills or claimed to have used AI when they had not, versus 15% of millennials, 10% of baby boomers, and just 6% of Gen X. Even more Zoomers (28%) avoided asking for help at work, and 23% avoided setting any limits on their workload, compared with 6% of boomers. Asking for help and setting limits are how workers keep a job sustainable and avoid taking on a workload that can lead to burnout. Gen Z is giving up both tactics and risks carrying these habits into the later years of their career.

Workers Are Applying Their Way Out

Workers usually job hunt for better pay or a better title. This year, many are doing it to get out of AI’s way, only to find AI screening the applications on the other end. Modern workers are caught in a feedback loop, where they need AI tools just to find a job that won’t rely on them.

Over half of workers (51%) applied for or seriously considered a new job for an AI-related reason in the past year. The generational spread is the widest in the study: 77% of Gen Z, 47% of millennials, 36% of Gen X, and 27% of boomers. For the youngest workers, AI is no longer a workplace condition to manage. It is a reason to leave.

Nearly half of workers (45%) applied for or considered a job because it appeared to offer greater job security, and 35% because the role seemed less likely to be affected by automation. Another 31% were drawn by an employer offering more AI training or guidance, helping workers stay competitive long-term.

More striking is the 36% who applied for or considered a role specifically because it required less AI use. Among Gen Z, that figure reached 58%. Even though the generation of digital natives is most fluent with AI tools, they’re the most opposed to using them. These early professionals may want to create work that speaks for itself, or Gen Z may be especially worried about the AI bubble popping and seek jobs with better longevity.

But even the job application process can stoke AI-related worries. Nearly a quarter of workers (24%) submitted an application without knowing whether a person would review it, and 19% never heard from a human at all. Another 14% said they received an automated rejection shortly after applying, likely due to an AI-powered selection tool. Faced with that, 18% rewrote their resume specifically to pass automated screening, rising to 31% among Gen Z. Others used AI to beat the AI: 14% used an AI tool to write or revise their application materials, and another 7% used AI to fabricate or exaggerate their skills. A minority (8%) withdrew from an application entirely because the process felt overly automated.

AI now influences every stage of the job process, from applications to interviews and beyond. Workers who are looking for an escape from workplace AI mandates often find they have to use the tool just for a chance to get away.

Why AI Anxiety Goes Untreated

None of these symptoms look like a reason to call a doctor. Trouble sleeping is a bad week. Answering messages at 10 p.m. is being reliable. Overstating your AI skills on a resume is what everyone does now. Because no single behavior is alarming on its own, workers keep explaining each one away — and a worry that never gets named as a problem never gets treated as one.

That matters most for the workers at the start of a career, because habits set early tend to stay set. A Gen Zer who has stopped asking for help and stopped declining assignments has removed the two things that make heavy workloads survivable, and they are removing them in their first years of work rather than their twentieth. Employers can change the arithmetic here at almost no cost: state plainly what AI will and will not change about a role, and pay for the training instead of letting people buy it on weekends.

What makes this a mental health story rather than a labor story is that the symptoms are treatable. Poor sleep, eroded confidence, and anxiety that follows someone home respond well to care, and they respond better the earlier someone names them. The workers in this study describe an ordinary, understandable response to a workplace changing faster than anyone expected—which is precisely the kind of thing that gets easier to carry once it stops being carried alone.