Universities across the United States are building AI classes, minors, and graduation requirements aimed squarely at students who have nothing to do with computer science. Purdue now requires every undergraduate to take an AI course. Ohio State has an AI fluency requirement with hands on workshops. Harvard is teaching freshman writing students how large language models work. Virginia Commonwealth University and Northwestern have added AI minors, and Northwestern is turning its into a full major.
This is happening at the same time computer science itself is shrinking. Enrollment in computer and information science programs at four year schools dropped more than 8 percent this past year, the first such decline in nearly two decades, according to the National Student Clearinghouse Research Center. A Stanford economist tracking the same trend called it the first drop in roughly twenty years.
The two trends are connected. Job postings for software developers have fallen sharply, down more than a third from pre pandemic levels according to Indeed Hiring Lab. A Stanford study using payroll data from ADP found that workers aged 22 to 25 in the most AI exposed jobs, including software development, saw employment drop by double digits since ChatGPT launched. Students are reading these numbers and deciding a computer science degree alone is no longer the safe bet it used to be. So they are picking up AI skills as an addition to a different major instead of betting the whole degree on coding.
Employers are reinforcing this shift from the other side. A recent survey by the National Association of Colleges and Employers found that more than one third of entry level jobs now require AI skills, a share that has nearly tripled in less than a year. That is not one third of tech jobs. That is one third of entry level jobs across all fields, including the marketing, insurance, and operations roles that fill up traditional companies.
This matters for anyone running a business, not just for parents of college students. The pool of new hires you draw from is changing shape. In two or three years, an accounting graduate or a marketing graduate is likely to walk in already comfortable prompting AI tools and questioning their output, because that is now baked into general education, not something they have to teach themselves after graduation.
There is a real cost buried in this trend too. A computer science education researcher at the University of Washington points out that when AI writes the code, it is also doing the thinking behind the code, which means some graduates may pick up the AI shortcut without ever learning the underlying logic. The same risk applies inside your own company. If you let AI handle a task end to end without anyone learning the fundamentals, you save time today but you may end up with a workforce that cannot catch mistakes or explain decisions when the AI gets something wrong.
The practical move is to treat AI literacy the way these schools are treating it: not as a specialist skill for the IT department, but as a basic expectation for every new hire and every existing employee, regardless of their job title.