Nearly Half of Gen Z Workers Say AI Is Making Them Dumber in 2026

A Generation That Already Suspects the Tools Are Costing Them Something

A new wave of research from GoTo and Workplace Intelligence covered by Fast Company surveyed 2,500 global employees and IT leaders and surfaced a quietly damning number: 46 percent of Gen Z workers believe their overreliance on AI tools is actively eroding their skills and making them less intelligent. The number for the broader workforce is 39 percent, which is bad enough, but the Gen Z spike is the data point that should make every hiring manager and every twenty-something rethink their daily workflow.

This is not a generation that hates AI. This is a generation that grew up with autocomplete, that used ChatGPT in college, that watched the North Dakota workforce conversation about AI job displacement risk play out in real time, and that has now spent two or three full years using large language models for everything from email replies to first drafts of code. They are the closest thing the labor market has to a control group for what heavy AI use does to a working brain, and they are telling researchers that something feels wrong.

The full GoTo numbers are worth sitting with. Fifty percent of employees across all age groups say they rely on AI too much. Thirty percent say they can no longer function without it. And sixty percent feel pressured to use AI tools to boost productivity regardless of whether the task actually calls for one. That last stat is the one that explains the others. When the pressure to use AI is constant and indiscriminate, the muscle memory you build is the muscle memory of asking, not the muscle memory of thinking.

What "Cognitive Offloading" Actually Means for Your Career

The term cognitive offloading has been kicking around academic psychology for more than a decade, but it has never been more relevant than it is right now. The short version: cognitive offloading is what happens when you outsource a mental task to an external tool, whether that tool is a calculator, a search engine, a GPS, or a chatbot. The longer version is more uncomfortable. Every time you offload a task, you also offload the neural rehearsal that would have built or maintained the skill behind it.

Researchers at institutions like MIT's Center for Brains, Minds and Machines have spent years studying how heavily we lean on external memory aids and what happens to recall, problem solving, and pattern recognition when we do. The pattern is consistent across studies: offloaded tasks get done faster in the moment, but the underlying competence atrophies over time, and the user typically does not notice the atrophy until they are forced to perform the task without the tool.

For a Gen Z worker who started using AI as a first-draft engine in their first job, this is not a hypothetical risk. It is the daily experience of opening a blank document, freezing, and reaching for the chatbot. The freeze is the warning sign. The freeze is the thing the GoTo respondents are reporting when they say AI is making them dumber. They are not getting dumber in any clinical sense. They are losing the on-ramp into their own thinking, which feels like the same thing.

Why Gen Z Is Feeling This Harder Than Everyone Else

Three things stack up to make this generation the canary in the AI coal mine.

First, Gen Z entered the workforce after generative AI was already mainstream. A millennial who learned to write reports the slow way in 2014 still has that scaffolding to fall back on. A 2024 graduate may never have written a long-form business email without AI assistance, which means there is no slow-way version of the skill stored in their muscle memory. The same dynamic is playing out for junior coders, junior analysts, and entry-level marketers in industries from healthcare to logistics, and it is starting to reshape how companies think about entry-level roles, as we covered in General Motors' AI-driven job cuts.

Second, Gen Z is under more productivity pressure than any generation that came before them at the same career stage. The sixty percent who feel pressured to use AI regardless of the task are disproportionately younger workers. Their managers, often millennials and Gen X, were sold AI as a productivity multiplier and are now measuring output against an inflated baseline. If the AI-assisted version of a task is the expectation, the un-assisted version no longer fits in the workday.

Third, Gen Z is the most metacognitive working generation we have ever had. They were raised on think pieces about screen time, dopamine, and attention, and they are quick to notice and name a cognitive cost. The 46 percent number is not necessarily evidence that they are more impaired than older workers. It may be evidence that they are more honest about the impairment. Either way, the self-report is useful, because it is the first step toward changing the habit.

The Skills Most at Risk of Atrophy

Not every task you offload to AI matters equally. Some are genuinely fine to hand off forever. Others are load-bearing skills that show up in promotion decisions, in interviews, and in the kind of work that gets you noticed.

The skills most at risk of quiet atrophy in 2026 are the ones that LLMs do passably well: first-draft writing, summarization, basic data interpretation, code boilerplate, meeting note synthesis, and the early structuring of a problem. These are also, not coincidentally, the skills that hiring managers look for in candidates moving from junior to mid-level roles. If you have outsourced all of them for two years, you may interview well using AI-assisted prep and then struggle in the role itself, a pattern showing up in feedback we have heard from teams covered in pieces like the senior accountant exodus.

Reading comprehension is another quiet casualty. When you ask an AI to summarize a long article or a research paper, you skip the part of reading where you wrestle with the argument, notice what is missing, and form your own opinion. Over months, the habit of skipping that wrestle shows up as a thinner point of view in meetings, which is the kind of thing that gatekeeps senior roles. The same pattern is true for any kind of analysis where the value is in the work of thinking, not the artifact at the end.

What Smart Gen Z Workers Are Doing Differently

There is a small but visible group of Gen Z workers who are using AI heavily and still building serious skill. They are not doing anything mystical. They are doing three things on purpose.

They write the first draft themselves, then use AI as a critic. Reversing the default order matters more than any other single habit. When you draft first and then ask the model to poke holes, you do the cognitive work of forming the argument and you get the benefit of a second pass. When you ask the model to draft and then you edit, you are doing the cognitive work of editing, which is a thinner skill.

They keep one weekly "no-AI" working block. Two to three hours, one day a week, where they research, write, or build without any AI assistance. The point is not productivity. The point is keeping the muscle warm. This is the same logic behind keeping a career resilience habit during a recession, which is to maintain the skill before you need it, not after.

They learn one technical skill the slow way per quarter. SQL, financial modeling, a new language, design fundamentals, something that requires real practice and cannot be faked with a chatbot at the moment of truth. The slow-skill habit is what separates the workers who will own AI-augmented roles in 2030 from the workers who will be replaced by the next model release.

What Managers Hiring Gen Z Should Actually Test For

If you are hiring a junior worker in 2026, the take-home assignment is no longer a useful filter. Any candidate with access to a modern AI tool can produce a polished take-home in an hour. The signal you actually need is whether they can think under live conditions.

The best Gen Z interview process in 2026 looks like this. A short async portfolio review. A live problem-solving conversation with no laptop open, where the candidate explains how they would approach a real problem at your company. A paired working session where the candidate uses AI in front of you and narrates the choices they are making. The narration is the test. A candidate who can explain why they trusted the model on one task and why they overrode it on another is a candidate who has metacognition about their own AI use, which is the rarest and most valuable trait you can hire for right now.

This is also why employers covered in pieces like the IBM biased hiring lawsuit and the broader protecting performers from AI threats conversations have started rethinking what entry-level even means. If the AI does the work a 22-year-old used to do, the 22-year-old you hire needs to be doing something different from day one, and you need to be able to identify which 22-year-olds can do that work.

The Pressure Problem Is a Management Problem

It is easy to read the GoTo numbers and put the burden entirely on workers. Use AI more thoughtfully. Take a no-AI block. Read the long article. All true, all useful, and all incomplete, because the 60 percent of employees who feel pressured to use AI regardless of the task are not making that pressure up. It is coming from somewhere.

It is coming from leadership decks that promised AI-driven productivity gains to the board. It is coming from middle managers who are graded on those gains and now have to show them. It is coming from team norms where the fastest reply in the channel is the AI-assisted reply, and the slower, more considered human reply looks lazy by comparison. It is coming from performance reviews that quietly assume an AI-augmented baseline without ever saying so out loud.

If you are a manager reading this, the work in front of you is to make it explicit. Tell your team which tasks they should use AI for and which tasks they should not. Build "thinking time" into sprint planning the way you build in code review. Stop rewarding speed-of-reply over quality-of-thought in your one-on-ones. The companies that figure this out in 2026 will be the ones whose junior employees actually grow into mid-level employees. The companies that do not will keep paying entry-level salaries for AI-assisted output and wondering why nobody on the team can run a meeting.

The Long View on This Generation

The 46 percent stat is going to get used as ammunition by every side of the AI debate. The doomers will say it proves the technology is rotting young brains. The boosters will say it is just adjustment anxiety and the next generation will be fine. The truth is closer to neither.

Tools change skills. They always have. Calculators changed what arithmetic fluency was worth. Spell-check changed what spelling fluency was worth. GPS changed what map reading was worth. In every case, the underlying skill did not disappear, it just got revalued, and the people who held onto a thin version of it on purpose ended up disproportionately valuable in the few situations where the tool failed. AI is going to do the same thing to drafting, summarization, analysis, and a dozen other knowledge-work skills.

Gen Z is the first generation to feel the revaluation in real time, in their own brains, while they are still trying to build a career on top of it. The fact that nearly half of them are already worried about it is, perversely, the most hopeful thing in the GoTo report. The workers who notice the cost are the workers who will do something about it. The workers who do something about it are the workers who will still be employable in 2030.

People Also Asked

Q: Is AI actually making Gen Z dumber, or is this just a perception thing?

A: It is not a clinical decline in intelligence, but the underlying mechanism is real. Cognitive offloading research from MIT, Stanford, and others has consistently shown that outsourced tasks lead to skill atrophy in the offloaded domain. Gen Z is reporting the feeling first because they offload more and notice it faster, and the 46 percent self-report in the GoTo data is best read as an early warning, not a diagnosis.

Q: How much AI use is too much for a Gen Z worker who wants to keep growing?

A: There is no single number, but the heuristic that works is "AI second, brain first." Draft the email, the analysis, or the code outline yourself first, then use AI to critique or extend it. If you cannot remember the last time you started a task without opening a chatbot, you are almost certainly past the line.

Q: Should Gen Z workers hide their AI use from their managers in 2026?

A: No. The trend is in the opposite direction. Managers are getting better at spotting AI-assisted work and what they actually want to see is judgment about when to use it. Narrating your AI choices in real time, in a meeting or a pull request, is a stronger career move in 2026 than pretending you did it all yourself.

Related Articles

• North Dakota Workers Face AI Job Displacement Risk
• General Motors Job Cuts Driven by AI
• Protecting Video Game Performers from AI Threats
• IBM Biased Hiring Lawsuit
• The Senior Accountant Exodus Movement
• Unemployment Insurance Preparedness for a Recession
• Tesla Robot Training Job Listing
• Software Engineer, Machine Learning at Whatnot
• DeFi Innovation Engineer Internship at Keyrock
• South Carolina Career Readiness Initiative

Future-proof your career. Metaintro tracks how AI is reshaping the job market and delivers career intelligence to your inbox weekly. Sign up free.