Why AI Anxiety Runs Deeper Than Your Job (and What's Actually Yours to Worry About)

If you've felt something closer to dread than curiosity about AI lately, you're far from alone, and the data backs that up in a way that might actually help.

A Gallup survey released in April 2026 found that anger toward AI among Gen Z rose from 22 to 31 percent in a single year, anxiety is holding steady at 42 percent, and nearly half of young workers now say AI's risks outweigh its benefits, up from 37 percent just a year earlier. This isn't a fringe reaction. It's a widespread, measurable shift in how an entire generation feels about the technology reshaping their working lives.

It isn't only Gen Z, either.

Millennials, now roughly 30 to 45 years old and often deep into building the career or business they've spent a decade on, are showing a quieter but genuinely real version of the same unease. Over a quarter of Millennials say they know they should be building new skills in response to AI but currently aren't, according to 2026 research from Careerminds, more than any other generation except Gen X.

Mercer's 2026 Global Talent Trends survey, covering 12,000 workers and leaders worldwide, found that fear of losing a job to AI has climbed from 28 to 40 percent across the workforce in just two years. For a lot of people in their late 30s and 40s, the feeling isn't loud anger. It's a low, steady hum of "I know I should be doing something about this, and I'm not sure what."

I want to take that feeling seriously rather than talk you out of it. But I also want to name something that I think gets lost in most conversations about AI anxiety: there are actually two very different fears tangled together here, and separating them matters.

Two Very Different Kinds of AI Anxiety

The first fear is personal: will AI take my job, or make the career I'm building obsolete. This fear is real, it's grounded in genuine data, and it's also something you have real agency over, through the skills you build, the work you choose to do, and the decisions you make about your own career.

The second fear is much bigger, and much less personal: AI being woven into systems that could cause serious harm at scale if something goes wrong. That covers real ground, from the most catastrophic end, nuclear command systems, autonomous weapons, critical infrastructure, to quieter but still serious concerns like algorithmic bias in hiring, lending, and criminal justice, and the sheer amount of personal data AI systems are trained on and collect.

This is a real and serious category of risk that credentialed researchers, policy experts, regulators, and technologists are actively working on right now. It is also not something any one person can personally fix through their own career choices, and I think conflating it with job anxiety often makes people feel a vague, unfocused dread about everything at once, rather than clear-eyed concern about the specific things that are actually theirs to think about. This piece focuses on the highest-stakes end of that spectrum, nuclear and military systems, since that's often the version of this fear that feels the heaviest, but the same underlying distinction applies to the rest of it too.

What the Data Actually Shows About Your Job

Economist Daron Acemoglu, who won the 2024 Nobel Prize for his research on exactly this question, estimates AI's effect on GDP over the next decade at a modest 1.1 to 1.6 percent, considerably more conservative than some of the bigger claims circulating in the media. That doesn't mean nothing is changing. Specific kinds of entry-level work, first-draft content, basic customer support, routine research, are already measurably affected. But "specific roles are shifting" is a very different, and much more actionable, statement than "everything is about to collapse."

It's also only half the picture. The World Economic Forum's Future of Jobs Report 2025 projects 170 million new roles created globally by 2030, against 92 million displaced, a net gain of 78 million jobs. Some of the fastest-growing specific titles are already visible right now: AI Engineer, currently LinkedIn's fastest-growing job title in the U.S., with postings up 143 percent year over year; MLOps Engineer, the role that keeps deployed AI systems running reliably, which LinkedIn's own data shows growing nearly tenfold over five years; AI Integration Specialist and Automation Architect, roles focused on connecting AI tools to a company's existing systems and workflows; Data Annotator, the human work of labeling and correcting the data AI systems learn from; and Chief AI Officer, a title roughly 60 percent of larger companies now either have filled or are actively hiring for.

None of these existed as common job titles five years ago. The pattern worth noticing isn't that work disappears. It's that it relocates, often toward roles that require exactly the kind of judgment and oversight a person, not a model, actually provides.

The workers navigating this with the most genuine confidence right now aren't the ones ignoring the shift. According to 2026 research from Careerminds, Gen Z workers are the generation most actively building new skills specifically because of AI concerns, at nearly double the rate of Millennials. Concern channeled into a specific skill or a specific plan tends to feel very different from concern with nowhere to go.

What's Actually Being Done About the Bigger Risks

Here's what I think doesn't get said enough in conversations about AI and catastrophic risk: serious people are already working on this, and there are real, concrete commitments already in place, not just hope.

In November 2024, the United States and China reached an explicit agreement that AI must never replace human judgment in decisions to authorize or launch nuclear weapons. U.S. Strategic Command's own commander testified to the Senate Armed Services Committee in 2025 that "AI will remain subordinate to the authority and accountability vested in humans" in the nuclear enterprise.

The United Nations passed its first resolution specifically addressing AI in nuclear command and control in late 2025. None of that means the risk is zero, or that the work is finished. Researchers at institutions like the Federation of American Scientists and the Stockholm International Peace Research Institute have pointed out, accurately, that investment in governance and oversight is still lagging behind investment in AI capability itself, and that more subtle risks, like AI-generated situational reports creating false confidence during a crisis, deserve continued, serious attention.

That's a genuinely different picture than "nobody is paying attention." It's closer to: this is being actively debated, legislated, and studied by people whose entire job is to get it right, and the foundational commitment, that a human being makes the final call, is already real and already in place.

What's Actually Yours to Carry

You're allowed to care about both of these things. But they don't require the same response from you.

The first deserves your attention and your action: a skill you build, a way you position your work, a plan you make. A concrete place to start is naming, specifically, which part of your own role leans hardest on the kind of work AI already does reasonably well, and which part leans on judgment, relationships, or taste that's genuinely yours, then deliberately spending more of your time and development in the second category.

That's a plan, not a vague resolution to "keep up," and a plan is usually what turns this particular fear into something workable.

The second, the bigger, systemic risks, deserves your awareness, and it's reasonable to stay informed, but it was never meant to be carried alone, at an individual level, as a source of daily dread.

If a general sense of AI dread has been sitting heavily on you, in a way that's affecting your sleep, your mood, or your ability to focus on your own work, that's worth taking seriously, and it's worth talking to someone, whether that's a coach, a therapist, or simply a trusted person in your life. That reaction is real. It just doesn't mean you're personally responsible for solving it alone.

In my experience as a life coach, the clients who move through this most steadily are the ones who get specific: naming exactly what part of this is theirs to act on, and consciously setting down the part that was never theirs to carry in the first place.

Feeling the Weight of This More Than You'd Like To?

If AI anxiety has been sitting heavier than it should, you don't have to sort through it alone. Book a Free 30-Minute Chat with Asi Efros to talk through what's actually yours to act on, and what isn't 👉 Book your free 30-minute chat here

I hope this blog post "Why AI Anxiety Runs Deeper Than Your Job (and What's Actually Yours to Worry About)" was helpful. You can find more insights like this on the blog 👉 Explore more insights on the blog

CREATE YOUR DREAM LIFE, CAREER & BUSINESS — one aligned step at a time