AI is no longer just a workplace issue for college students. It is beginning to influence what they study, where they want to work, which skills they develop, and even whether they believe their degrees will be worth the investment.
According to a CNBC and SurveyMonkey survey of 1,686 U.S. students and workers enrolled in school, nearly 40% have considered changing their major, field of study or coursework because of AI. Among students who are not currently employed, that figure rises to 44%.
The concern is broader than majors. Nearly one-third (31%) said they have changed or considered changing their target industry because of AI, while 45% have reconsidered the skills they are developing. About 34% have reconsidered which jobs they apply for, and 35% have reconsidered the companies they want to work for.
Those numbers point to a significant shift in how young people are thinking about education: students are increasingly treating AI disruption as a factor in deciding what kind of future they should prepare for.
But there is an important problem with that logic.
Students are making decisions about a labor market that hasn't finished changing
The anxiety is understandable.
A student spending four years and potentially tens of thousands of dollars on a degree wants some confidence that there will be a viable career waiting at the end. AI has made that calculation considerably harder.
Stephanie Park, a 20-year-old liberal-arts student who hopes to become a doctor and researcher, is already wondering what those professions will look like when she finishes her training.
Her concern isn't simply that AI will eliminate doctors. It's that AI could change the relationship between doctors and patients.
As people increasingly turn to chatbots for information, advice and even emotional support, Park worries that the human side of medicine could receive less attention.
That uncertainty is spreading across other fields as well.
Allison Shrivastava, an education and labor-market economist at Niche, says students are hearing contradictory predictions: AI will create enormous numbers of jobs; AI will eliminate enormous numbers of jobs.
The danger is that students may make permanent educational decisions based on a labor market that is still in flux.
The uncomfortable truth is that nobody knows exactly which occupations will be most transformed by AI five or ten years from now.
That makes choosing a major based primarily on AI anxiety a risky strategy.
Computer science is the obvious warning sign
Computer science illustrates the problem particularly well.
For years, students were told that learning to code was one of the safest bets they could make. The technology industry expanded rapidly, salaries were high, and demand for technical workers appeared almost limitless.
Then the market changed.
The unemployment rate for computer science graduates reached 7% in 2024, according to data cited by CNBC, while major technology companies pulled back from the aggressive recruiting that characterized the previous hiring boom.
For a student watching headlines about AI coding tools, the conclusion might seem obvious:
Why study computer science if AI can write code?
But that conclusion is too simplistic.
Computer science isn't synonymous with working for a major technology company. Banks, hospitals, manufacturers, governments and virtually every large organization increasingly depend on software and computational systems.
The more useful question isn't whether AI will reduce demand for certain coding tasks.
It's whether someone with strong technical knowledge can become more valuable by knowing how to use, evaluate and direct AI systems.
That distinction matters.
The safer bet may be skills, not majors
Career coach Jessica Roffe argues that students should focus less on finding the supposedly “AI-proof” major and more on developing skills that remain valuable as technology changes.
That means building capabilities such as:
Critical thinking
Problem-solving
Curiosity
Skepticism
Communication
The ability to evaluate AI-generated information
The last one may become particularly important.
AI can generate an answer quickly. That doesn't make the answer correct.
As AI becomes embedded in more professional workflows, employers may increasingly need people who can determine when an AI-generated result is useful, when it is wrong, and when the underlying question has been misunderstood altogether.
In other words, AI literacy may matter less as a standalone technical skill than as a layer added to an existing area of expertise.
A nurse who understands AI, a lawyer who understands AI, an engineer who understands AI or a financial analyst who understands AI may all have an advantage over someone who possesses only generic AI familiarity.
Don't confuse uncertainty with obsolescence
The survey reveals something important about students' psychology, but it doesn't establish that particular degrees are becoming bad investments.
Nearly 40% reconsidering their major is evidence of uncertainty, not evidence that 40% of majors are doomed.
That's a crucial distinction.
Students are being exposed to extraordinary claims about AI while simultaneously trying to make decisions with long-term consequences. It is reasonable for them to reassess their plans.
But abandoning a field simply because AI might change it could be exactly the wrong response.
Medicine will change. Law will change. Finance will change. Software development will change. Education will change.
That doesn't necessarily mean those professions disappear.
More likely, the tasks inside them change.
The students best positioned for that future may therefore not be the ones who successfully predict which jobs AI will eliminate.
They may be the ones who develop deep expertise in something they care about while becoming exceptionally good at working with new technology.
The bigger lesson
Students shouldn't ignore AI when choosing a major.
But they shouldn't let AI headlines choose their major for them either.
A better strategy is to ask three questions:
What am I genuinely interested in?
What am I good at—or willing to become good at?
How can I combine that expertise with increasingly capable AI tools?
The labor market of 2030 cannot be predicted with precision from the labor market of 2024, just as today's market could not have been perfectly predicted from the technology landscape a decade ago.
Students don't need an “AI-proof” career.
They need adaptable expertise.
And that may ultimately be more valuable than trying to guess which profession AI will—or won't—disrupt.