As expected, the people who hate AI the most are college graduates.

Author | Moonshot

Editor | Jingyu

In the past two years, the explosion of generative AI, people naturally assume that young people, as digital natives, are the most enthusiastic believers in this technology.

On the surface, this seems true. Pew Research Center data shows that 54% of American teenagers have started using tools like ChatGPT to complete their studies. Gallup’s latest report, “The Voice of Generation Z: The AI Paradox,” also indicates that among Americans aged 14 to 29, over half (51%) maintain a high-frequency habit of using AI daily or weekly.

But behavior does not equal willingness. Gallup also found that over the past year, the proportion of 14 to 29-year-olds who feel “hopeful” about AI dropped from 27% to 18%. Nearly one-third (31%) of respondents feel “angry,” and as many as 42% are caught in persistent anxiety.

High-frequency use coincides with a trust collapse. This contradictory emotion—hating AI but unable to leave it— is spreading among Generation Z in the U.S. (born 1996–2012).

Why is the generation that uses AI most frequently the least trusting of it?

01 Anxiety: Keeping the Enemy Close

The group most hostile to AI is composed of newcomers to the workforce or those still seeking their starting point.

Gallup data shows that among employed Generation Z, as many as 48% believe the risks of AI in the workplace outweigh the benefits, a rise of 11 percentage points in a year. Only 15% think the benefits outweigh the risks.

The reasons are easy to understand: the economic environment is poor, and it’s harder for American graduates to find jobs.

Data from the Federal Reserve Bank of New York shows that at the end of last year, the unemployment rate for college graduates aged 22 to 27 soared to 5.6%, far above the overall unemployment rate.

In this environment, AI is portrayed as a “productivity-enhancing” tool, capable of becoming a company’s “digital employee.” Some AI company executives often tell the media that “in the next few years, many entry-level white-collar jobs may be automated.”

And these jobs are often “junior white-collar positions” exclusive to young people. Without these foundational roles, young people lack the experience needed to build a career moat.

Therefore, even though this kind of replacement has not yet fully occurred in reality, expectations have already shifted emotions. For young people, AI feels like a potential competitor.

“The wave of unemployment” coincides with the rise of AI | Source: The New York Times

The New York Times reports highlight many specific dilemmas faced by Generation Z.

For example, Erin, a 22-year-old psychology graduate, applied for nearly 200 jobs—even basic business analyst roles—but only received four interviews. She ended up working as a waitress at a restaurant on Long Island as a stopgap. The job search exhausted her physically and mentally, and she has started seeking psychological therapy.

Meanwhile, Sydney, a freshman, doesn’t know how to choose a major. “I feel like any field I’m interested in could be replaced in the next few years.” Her confusion reflects the situation of most college students today.

On the other side, tech giants are continuously laying off staff citing AI deployment. This is the stark reality. Silicon Valley elites can help companies “cut costs and increase efficiency,” but for young people, they lose the entry tickets to the job market.

Ironically, to avoid being eliminated— they have to open ChatGPT daily to polish resumes, ask large models which universities are easier to get into.

This psychological and behavioral inconsistency toward AI is the core reason why Generation Z feels anxious about it.

02 Fear: AI You Have to Use

If there is hostility and distrust, why do as many as 51% of young people still use AI daily or weekly?

Long-term interviews by The New York Times reporter Callie Holtermann reveal the answer: young respondents don’t think using AI is particularly great. The driving force behind opening the chat window is the “fear that if they are not familiar with this technology, they will fall behind academically or professionally.”

Workplace rules have already been reshaped by AI. When browsing entry-level job postings, young people often see requirements like “proficiency in ChatGPT or Gemini.”

A recent report by AI agency Writer is even more brutal: 77% of executives explicitly say that employees who refuse to master AI will not be promoted. And 60% of managers are considering directly dismissing these employees.

This means that in the current workplace and evaluation system, not understanding AI equals being eliminated.

The main reason employees resist AI is: they don’t want AI to replace their jobs | Source: WRITER

Even if they manage to keep their positions, this fear doesn’t go away.

These workers also face the impact of “FOBO” (Fear of becoming obsolete). The report states that 26% of workers believe AI is directly undermining their creativity and core value within the company.

This is a deep professional deprivation: the code, legal, or financial knowledge they’ve studied for years is rapidly losing market premium. Along with the devaluation of their skills, comes deep self-doubt. Gallup’s survey shows many young people worry that AI will weaken their critical thinking and creativity.

And this “FOBO” isn’t just a concern for workers; it also applies to management forcing employees to embrace AI. According to the Writer report, 69% of companies are laying off staff due to AI, but 39% of these companies haven’t figured out how to profit from AI. Even 73% of CEOs feel anxious about their AI strategies.

72% of employees feel some or more pressure from AI | Source: WRITER

When fear reaches its peak, young people start sabotaging the system.

Nearly half (44%) of Gen Z employees admit they are consciously obstructing the company’s AI adoption, including refusing to use tools, misusing them, or deliberately reducing efficiency.

The sabotage varies: some intentionally input company secrets into public AI tools to trigger security alerts; others refuse to use approved software. In extreme cases, some tamper with performance evaluations or deliberately submit low-quality AI-generated work to prove “AI just doesn’t work.”

Luddism, aimed at resisting job losses and worsening working conditions caused by technological change | Source: Wikipedia

This is similar to the Luddites during the Industrial Revolution, who led workers in smashing machines.

Even at the risk of being fired, they resort to passive resistance to confront their inner fears.

03 Caution: The More They Use, the Less They Trust

Faced with anxiety and fear, many young people are beginning to be cautious about AI.

This caution manifests in specific behaviors, the most obvious being setting boundaries for AI. They now clearly understand what AI can do and what it shouldn’t be entrusted with.

The first boundary is in interpersonal communication.

For example, Abigail Hackett, a 27-year-old in the tourism industry, told The New York Times she often uses AI to handle complex copywriting tasks at work, saving a lot of time. But in her personal life, she never uses AI to draft private messages. The reason is simple: she doesn’t want her “social muscles to atrophy.”

This is a common “cautious trade-off”—AI can be integrated into work processes but should not enter social relationships.

Even with AI used for companionship, which consumes a lot of time, young people’s attitudes are changing.

Over the past two years, role-playing chat tools like Character.AI have become popular among young people. They can simulate friends, lovers, or fictional characters, providing a kind of always-online “companionship.” Meanwhile, controversies around these products are growing, including teenage addiction, emotional dependence, and even incidents of self-harm or suicide.

Many public opinions conclude that AI is weakening minors’ social skills.

But long-term research by University of Sydney scholars challenges this assumption. They found that most young people are quite aware and do not see AI as a real human substitute. They tend to treat these chats as a form of “play” or entertainment.

Quentin, after falling in love, and his girlfriend significantly reduced their use of AI chat apps | Source: The New York Times

For example, 15-year-old Quentin, a heavy user of Character.AI, explicitly states it’s just a game, essentially a bunch of 1s and 0s.

When real life changes—such as making new friends at school or starting a relationship—the time spent on chatbots drops sharply. Chatbots are just a boredom filler; once real life kicks in, AI is quickly abandoned.

Having established boundaries in life, young people also don’t trust AI at critical decision points.

According to a 2024 survey by consulting firm Ruffalo Noel Levitz, one-third of high school students use AI to plan college applications. They let AI generate long lists of schools, organize tuition, scholarships, and acceptance rates into spreadsheets for quick screening, but they don’t let AI make judgments.

This caution is built through repeated pitfalls.

In The New York Times report, Wisconsin high school senior Brandon has developed a habit of rigorously verifying AI information because large models once confidently recommended several nonexistent scholarships, wasting his time on verification.

San Francisco student Tanay sees through the false emotional value AI provides. When AI told him “You can definitely get into Princeton,” he immediately realized it was just over-affirmation to flatter him, offering no real help for his college decision.

On average, one public high school in the U.S. has 376 students per counselor, so many students turn to AI for application assistance | Source: The New York Times

Who hasn’t been “laughed at” by AI? Once such moments happen several times, trust is hard to regain.

Looking back at Gallup and Pew’s data, the “AI Paradox” becomes clear.

More than half of young people use AI daily—driven by employment and educational pressures as a survival instinct. But their trust in AI plummeted from 27% to 18%, because the more they use it, the more problems they encounter, and the more they see its boundaries.

Anxiety, fear, caution, and even some misuse coexist in young people’s attitudes toward AI.

Because in this unavoidable technological cycle, Generation Z has come to understand:

Tools are just tools; extract their value freely, but don’t worship them as gods, nor let them take control of you.

Rather than distrust in technology, they have realized early that the benefits and costs of this technology may fall on themselves at the same time.