AI Is Quietly Erasing Your Confidence — and You Can Feel It

TL;DR

• The GoTo/Workplace Intelligence "Pulse of Work in 2026" study found AI saves workers 2.3 hours a day — but 50% say they rely on it too much, and 39% say it's making them less intelligent.
• Gen Z is the canary: 40% say they can't function without AI, 46% say it's eroding their skills, and 50% believe AI reliance will hurt their career prospects.
• This is not a technology story. It is a confidence story. The quiet cost of AI adoption is the slow erosion of self-trust — and the people who feel it most are the ones who've never worked without it.
• The fix is not to use less AI. It is to use it differently — as a thinking partner, not a thinking replacement.

What Happened

On May 19, GoTo and Workplace Intelligence released their "Pulse of Work in 2026" study, a global survey of knowledge workers examining how AI is reshaping the workplace. The headline finding was designed to travel: employees using AI save an average of 2.3 hours a day.

That number got the headlines. It should not have.

Buried beneath the productivity lede was a set of findings that, taken together, describe something more consequential than time savings. Half of all employees say they rely too much on AI. Thirty-nine percent say relying on it is making them less intelligent. And among Gen Z — the first generation to enter the workforce with AI tools already embedded in daily workflows — the numbers are sharper: 40% say they can't function without AI, 46% say AI is eroding their skills, and 50% believe that relying on AI will hurt their long-term career prospects. Thirty-five percent think AI is already doing their job better than they can.

"The productivity gains are real and measurable," GoTo CEO Rich Veldran told Newsweek, "but there's real concern that when not used properly, AI is also eroding skills, judgment and accountability." Then the line that should have been the headline: "Companies need to understand that productivity and capability are not the same thing."

What It Actually Means

This is not a technology story. It is a confidence story wearing a productivity costume.

The mechanism is straightforward, and it operates at the level of identity, not output. When you outsource a task to AI — writing an email, summarising a document, making a decision — you gain speed. But you also lose something: the small, repeated experience of doing the thing yourself and discovering that you can. Each time you reach for the AI instead of your own mind, you make a tiny withdrawal from the account of self-trust. Over months, the balance drops. You start to believe — not intellectually, but at the level of reflex — that you need the tool to function.

That is what 40% of Gen Z workers are reporting. They are not saying AI is convenient. They are saying they cannot function without it. The distinction matters.

The study's Gen Z findings are particularly significant because this cohort has no "before" to compare against. Older workers remember writing their own emails, forming their own arguments, making their own decisions. They can feel the difference between using AI as a tool and being used by it. Gen Z entered the workforce with AI already ambient. For many of them, the first draft of every email, every analysis, every difficult conversation was mediated by a model. They have never built the baseline confidence that comes from doing hard things without assistance.

And they know it. That is the most important finding in the study: Gen Z is not oblivious to what is happening. They are reporting it. Forty-six percent say AI is eroding their skills. Fifty percent say it will hurt their careers. They are watching themselves become dependent in real time, and they are scared.

The Hype Deconstruction

The productivity narrative — "AI saves 2.3 hours a day!" — is true but misleading. It frames AI adoption as a pure efficiency gain, the way email was faster than fax. But email did not make people feel less capable of communicating. AI, used in a particular way, does.

The productivity frame also obscures the distribution of the confidence cost. The workers who feel the erosion most acutely are the ones with the least professional experience — the people who most need to build self-trust. AI is not taking their jobs. It is taking their belief that they could do their jobs without it.

The Stakeholder Landscape

Gen Z workers are the primary affected group. They are gaining productivity and losing confidence simultaneously, and they are articulate about the trade-off. This is a generation that talks about mental health openly; they are applying the same vocabulary to their relationship with AI.

Managers and organisations face a quiet crisis. A team that produces more output but has less confidence in its own judgment is not a stronger team. It is a faster team with a brittle foundation. When the AI goes down, or when the situation requires genuinely novel thinking, the capability gap will surface.

Educators and universities are implicated. If students are graduating into a workforce where they immediately feel dependent on AI, the question is not just whether they learned AI skills — it is whether they built enough unassisted competence to trust themselves.

AI companies have an incentive to frame this as a training problem, not a product problem. "Use AI as a thinking partner, not a replacement," Microsoft's Work Trend Index advises. But the products are designed for speed, not for skill-building. The UX rewards outsourcing, not learning.

What This Means for You

If you are a knowledge worker: Run a one-week experiment. For tasks where the stakes are low, do the first draft yourself — without AI. Notice the discomfort. That discomfort is the sound of a muscle you have stopped using. The goal is not to abandon AI. It is to know, with evidence, that you can still do the thing without it. That knowledge is the foundation of confidence.

If you manage people: Stop measuring AI adoption. Start measuring AI dependency. Ask your team: "What would you still feel confident doing if the AI were unavailable tomorrow?" The answers will tell you more than any productivity dashboard.

If you are Gen Z entering the workforce: The fact that you are worried about this is actually a good sign. The people who should worry are the ones who have outsourced their thinking and stopped noticing. Build one area of unassisted competence — one skill, one domain, one type of task — where you know you don't need the AI. That becomes your anchor.

The Uncertainty Ledger

• The GoTo study is self-reported survey data, not a controlled experiment. Self-reports about skill erosion are subject to perception bias — people may feel less capable than they actually are.
• The Gen Z sample size and methodology are not fully disclosed in the public summary. The findings are directionally significant but should be treated as a signal, not a precision measurement.
• The counter-narrative — that AI is augmenting rather than replacing human capability — has its own evidence base. Microsoft's Work Trend Index found that 58% of AI users say they are producing work they could not have produced a year ago, rising to 80% among "Frontier Professionals." The truth is probably both: AI is simultaneously expanding capability and eroding confidence, and the net effect depends on how it is used.

Bottom Line

The productivity numbers are real, but they are not the story. The story is that a generation of workers is telling us, in their own words, that AI is making them feel less capable, less intelligent, and less confident in their own future. That is not a training problem. It is a design problem, a management problem, and — for anyone who cares about human capability — a warning. The fix is not to use less AI. It is to use it in a way that builds self-trust rather than eroding it. The difference is not in the tool. It is in the hand that holds it.

Sources:

• Tier 1: Newsweek — "AI Is Making Workers Faster. That May Be the Problem." (May 19, 2026)
• Tier 1: Newsweek — "Gen Z Says They Can't Function Without AI" (May 19, 2026)
• Tier 2: Forbes — "Microsoft Work Trend Index 2026 Shows AI Productivity Is Not Enough" (May 19, 2026)
• Tier 2: HR Dive — "Don't discount human skills or older workers in AI upskilling, expert warns" (May 19, 2026)