Why Some Young Adults Use AI—and Still Don’t Trust It

Someone can use AI every week and still worry about what it is doing to work, privacy or the quality of information. That is not automatically a contradiction. Using a tool, trusting its answers and supporting every way it is deployed are different decisions.

The tempting headline is that Gen Z and Millennials dislike AI while Gen X welcomes it. The evidence supports a more interesting question: what makes a person cautious, and what kind of AI use would earn their trust?

What recent surveys actually show

Pew Research Center’s June 2026 US survey found that 55% of adults under 30 were more concerned than excited about AI. The analysis says their concern was now similar to that among adults in their 30s and 40s and those 65 and older. Concern was not unique to young adults. [1]

A separate Pew workplace study published in February 2025 found that workers aged 18–29 were more likely than older groups to use AI chatbots at work at least a few times a month. [2] These are different surveys and dates, so they do not establish that the same individual both uses and distrusts AI. They do show why adoption and enthusiasm should not be treated as synonyms.

Both findings concern the United States. They are not Canadian population estimates, and an “under 30” category does not correspond neatly to every member of Gen Z. Age-group differences also do not prove a lasting generational personality.

Canadian use is not the same as Canadian trust

Statistics Canada’s March 2026 worker survey measures workplace use, including differences by age and occupational exposure. Among workers in high-exposure, high-complementarity occupations, use was more common in the 25–54 group than among younger or older workers. [3]

That classification concerns tasks and how AI might complement work. It does not tell us how much each generation trusts AI, nor why a particular person likes it. Job access, workplace rules and the tasks available to a worker can all affect opportunities to use it.

Five reasons worth investigating

The following are explanatory possibilities, not claims that the surveys proved each cause.

Career entry: someone trying to secure a first role may view automated junior tasks differently from an established worker who uses the same tools to save time. The question is whether employers provide meaningful learning opportunities, not whether every young person opposes technology.

Reliability: using AI exposes you to both useful output and confident mistakes. A cautious user may have learned to verify answers rather than reject the tool altogether.

Privacy: a person may welcome help drafting a document while objecting to uploading customer records or personal messages. Trust depends partly on control over what data is shared.

Authenticity: automated text, images and voices can make it harder to judge what reflects a real person’s work or experience. People may value disclosure without objecting to all AI-assisted creation.

Fairness and control: an employee might accept an optional assistant but dislike opaque hiring or performance decisions. Trust in a writing aid is different from trust in a system that affects someone’s livelihood.

Why Gen X can look more comfortable in some situations

Experience can provide context for checking output, and an established role may offer greater control over how tools are used. Those are plausible situational advantages, not evidence that everyone in Gen X is more trusting.

Other experienced workers may face retraining costs, privacy concerns or fear of displacement. Younger workers may be enthusiastic and skilled adopters. Income, occupation, education and individual experience can cut across generation labels.

A survey would need to ask comparable questions and account for those differences before supporting a firm explanation of why one generation trusts AI more.

What could earn trust at work?

Employers can make adoption more credible by explaining the actual task, what information is used, who checks the result and what happens when it is wrong. Workers need a way to question a decision and develop skills, rather than being told merely to become more productive.

A practical trial should measure accuracy and checking time alongside speed. “This saved ten minutes but introduced a serious error” is a different result from “this saved ten minutes and passed review.”

What readers can do

• Separate your concerns: accuracy, privacy, employment, ownership or accountability.
• Test a small task using fictional or permitted information.
• Record errors and the effort needed to check the result.
• Ask what human review and appeal exist when a system makes consequential decisions.
• Make your decision task by task; you do not have to either trust everything or refuse everything.

For Leaf, the useful conversation is less about which generation “gets it” and more about what people need before accepting a tool in their lives.

Research checked October 8, 2026. Survey findings are labelled by country and date. The proposed explanations are editorial interpretations to investigate, not demonstrated causal findings.