Younger workers feel confident using AI — but does that confidence actually match their results?
Do younger workers overestimate their AI skills more than older workers?
This explores whether younger workers are more likely than older workers to think they're good with AI when their results suggest otherwise. The corpus has one direct data point on age, plus a deeper body of work on why AI makes almost anyone overrate their own skill.
This explores whether younger workers are more likely than older workers to think they're good with AI when their results suggest otherwise. The short answer: one survey points that way, but the more interesting finding is that the mechanisms behind overconfidence don't depend on age at all. WalkMe's survey of about 2,000 US workers found that 90% felt confident using AI, yet only a quarter said it worked on the first try, and half said AI took them longer than doing the task by hand. The gap between confidence and results was widest among younger workers Why do workers feel confident with AI but get poor results?. Keep in mind what that survey measured. It compared people's confidence with their own reports of how things went. It never tested anyone's actual skill, so it can't tell us whether younger workers are worse with AI or just more willing to say they're confident.
That difference matters because self-assessment of AI skill turns out to be close to useless. A pooled analysis of three studies found a correlation of only .055 between how competent people said they were with generative AI and how they actually performed, and the confidence interval included zero Can self-ratings replace objective performance scores for AI competence?. If self-ratings barely track real ability for anyone, a survey built on self-ratings can show that one group is more confident. It can't show that the group is less skilled.
The broader research explains why the confidence gets inflated in the first place. Researchers call it the "LLM Fallacy": when AI output is smooth and polished, people start counting it as evidence of their own ability Do AI-assisted outputs fool users about their own skills?. Fluent text sends a misleading signal, because reading something that flows well feels like having understood or produced it Does processing ease mislead users about their own competence?. One analysis names four mechanisms that compound each other: it's unclear who did what, the output feels effortless, the thinking gets handed off to the tool, and the process stays hidden How do AI tools trick users into overestimating their own skills?. This error is separate from hallucination or automation bias. It can happen even when the AI's answer is correct, which is why making AI more accurate won't fix it How does AI-assisted work reshape how people see their own abilities?. Nothing in these mechanisms is specific to age. A plausible guess, which the corpus doesn't test, is that younger workers simply use AI more, and every extra use is another chance for the misattribution to build.
The usage angle leads somewhere unexpected. Anthropic found that the people who delegate the most work to Claude are also the most optimistic about their careers and the most likely to say their skills are gaining value. That's correlation within Anthropic's own users, with no independent skill check Does delegating work to AI actually damage worker skills?. Meanwhile, research on workplace risk warns that even "augmentation" use, where AI assists instead of replacing, can slowly erode the skills people would need to catch AI's mistakes Does AI augmentation protect workers from skill erosion?. Put together, heavy users may feel most confident at exactly the point where their unassisted skills are weakening.
The flip side is worth knowing. A hiring experiment found that listing AI skills partly offset the interview penalty older candidates usually face Can AI skills help older or less-educated job candidates?. Employers reward AI skill as a signal, yet that signal is mostly self-reported and, as shown above, barely related to real performance. People also expect to be judged as less competent when they admit using AI, so they often hide it Do people fear judgment when they use AI at work?. AI competence is therefore claimed loudly on résumés and hidden in daily work. That makes it hard to measure for any age group, and it means "younger workers overestimate more" is still a hypothesis that needs testing against real performance.
Sources 10 notes
WalkMe's survey of 2,037 US workers found 90% feel confident using AI, but only 25% report it works on first try and 50% spent more time using AI than doing tasks manually. The gap widened most among younger workers, suggesting overestimation of skill.
A pooled analysis of three studies found a correlation of only .055 between self-reported and objective measures of AI competence, with confidence intervals including zero. This provides no basis for substituting self-assessment for demonstrated performance.
Research identifies a systematic cognitive attribution error where individuals integrate AI-generated outputs into their capability identity, believing they possess skills they don't actually have. This occurs when task output is seamless and fluent, obscuring the human-AI boundary.
High-quality AI output triggers a metacognitive heuristic: users experience fluency as a signal of their own capability, even though they didn't generate it. This self-directed fluency illusion systematically inflates perceived competence because LLMs optimize for fluency regardless of user understanding.
Attribution ambiguity, fluency illusion, cognitive outsourcing, and pipeline opacity combine to systematically misattribute AI outputs as user competence. The effect is multiplicative—each mechanism amplifies the others.
Show all 10 sources
Research shows the LLM Fallacy operates through misattribution of AI outputs to personal capability, independent of output accuracy or reliance behavior. It requires interventions that clarify human-machine contribution boundaries, not just better system accuracy or forced verification.
Anthropic's Economic Index found survey respondents who delegate most work to Claude expect better career outcomes and report skills gaining value. However, the study shows only correlation within Anthropic's own user base, not causation or independent skill validation.
Research mapping 8,356 workplace AI risk scenarios found that augmentation mode does not inherently prevent harm. Overreliance on AI agents can gradually erode worker skills and their capacity to provide meaningful oversight, undermining augmentation's core safety justification.
A hiring experiment found that AI skills reduced interview invitation penalties for older candidates and those with associate degrees rather than bachelor's degrees. The effect was strongest for office assistant roles and weaker for graphic designers.
Across four experiments with 4,439 participants, people using AI expected others to judge them as less competent and diligent, and reported lower willingness to disclose AI use to managers and colleagues. The gap suggests a social cost that users foresee and act on.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- Evidence of a social evaluation penalty for using AI
- The LLM Fallacy: Misattribution in AI-Assisted Cognitive Workflows
- Beyond AI Literacy: A Structured Review and Exploratory Meta-Analysis of Measures for Competent Generative-AI Use
- AI Skills Improve Job Prospects: Causal Evidence from a Hiring Experiment
- Using AI More Does Not Reassure Workers, Managers Do
- What 81,000 people told us about the economics of AI
- Toward Measuring AI's Effects on Skill Formation: The Stock-Formation Gap
- Introducing Anthropic Interviewer: What 1,250 professionals told us about working with AI