AI Entry-Level Jobs Are Disappearing Faster Than Colleges Can Adapt

More career coverage. The first wave of artificial intelligence in the workplace did not eliminate jobs. It eliminated the first job most young people used to get.

For decades, the entry-level position was a clear stepping stone. A recent graduate joined a law firm as a paralegal, a bank as a junior analyst, a consultancy as a research associate. The work was repetitive but it taught the next generation of professionals how the industry actually worked.

According to a 2025 analysis from the Stanford Digital Economy Lab digitaleconomy.stanford.edu, early-career workers in fields exposed to large language models, including software engineering, marketing, customer support, and paralegal work, have seen the steepest declines in hiring since 2023. Workers aged 22 to 25 in those fields have experienced roughly 20 percent fewer job postings than would have been expected based on pre-AI trends, according to the same research.

The pattern is consistent across multiple studies. A 2024 paper from the Brookings Institution www.brookings.edu found that AI-exposed roles are seeing slower growth in entry-level postings than senior roles, and a separate analysis from the Federal Reserve Bank of New York www.newyorkfed.org showed that college graduates' unemployment rate has risen faster than the overall rate.

What the AI Tools Are Doing

The shift is being driven by tools that automate the kinds of tasks that used to be the job of new hires.

In software engineering, AI coding assistants like GitHub Copilot and Anthropic's Claude Code can write first-draft code, debug simple programs, and convert specifications into working software. Junior engineers at several major tech companies have told reporters that the share of their work that is actual coding has fallen sharply, while the share that is reviewing and editing AI-generated code has risen.

In marketing and advertising, AI tools now generate first-draft copy, design social media assets, and produce video. Junior copywriters and designers who used to spend their first two years producing high-volume work are now spending more time reviewing AI output and learning strategy.

In customer support, AI chatbots are handling a growing share of routine inquiries. Several large companies have publicly disclosed that they have not backfilled support positions that were vacated through attrition, because the AI tools can handle the volume that used to require additional headcount.

In legal work, junior associates at major law firms are doing fewer document reviews and more AI-supervised work. Several firms have told the Wall Street Journal that they are hiring smaller classes of associates than in past years.

The Counterargument

It is important to be careful about what the data actually shows.

The Stanford Digital Economy Lab paper finds that AI is having a measurable effect on entry-level hiring in some fields, but the same paper also finds that the effect is uneven across industries and geographies. AI exposure is correlated with slower entry-level hiring, but correlation is not causation. Some of the slowdown reflects broader trends in tech, finance, and media that began before AI tools were widely deployed. More career coverage

The Brookings analysis is more cautious. It documents that AI exposure is associated with changes in hiring patterns but stops short of claiming that AI is the only cause. The New York Fed research is similarly careful to note that the unemployment rate of college graduates is higher than in past years but not historically high by the standards of past recessions.

What the evidence does support is that the nature of entry-level work is changing. Even in fields where total headcount is stable, the share of work that is human-only is falling.

What This Means for Gen Z Workers

For Gen Z workers, AI entry-level jobs are reshaping the practical implications.

First, the kinds of jobs that used to be available straight out of college are not guaranteed. The resume, the cover letter, and the interview are not enough. Candidates need to demonstrate specific skills in working with AI tools, whether that is prompt engineering for a marketer, code review for a software engineer, or AI-supervised analysis for a finance role.

Second, the skills that matter are shifting. Knowledge of how to use a particular tool is less important than knowing how to evaluate the output of any tool. The same generalist skills that have always been valuable, critical thinking, written communication, the ability to work across teams, are becoming more important, not less.

Third, the career ladder is changing. If the entry-level job is being automated, the path to senior positions is no longer as clear as it was a generation ago. Several universities, including Stanford, MIT, and Carnegie Mellon, have updated their curricula in the past year to emphasize AI-augmented work and to add new courses on topics like AI ethics, model evaluation, and human-AI collaboration.

What Workers Should Do

The advice from labor economists and career coaches is consistent.

Build a portfolio, not just a resume. Show that you can do real work with AI tools, ideally on problems that the employer cares about.

Pick fields where AI is a tool, not a replacement. AI is unlikely to replace doctors, nurses, electricians, or therapists anytime soon. It is more likely to change the way those jobs are done.

Be ready to change jobs more often than previous generations did. The average tenure of a job in 2026 is shorter than it was in 2006, and Gen Z workers are already changing jobs more frequently than any previous generation at the same age.

Stay current with the tools. The most useful skill in 2026 is the ability to evaluate the output of an AI system, regardless of which system it is. The same skill will be useful in 2030, even as the systems themselves change.

The transition is real. For those affected by AI entry-level jobs, the right response is not to panic and not to ignore. It is to understand what is changing and to build the skills that will still matter when the next round of AI tools arrives.