For every 10‑percentage‑point increase in the share of automatable tasks associated with a given major, graduates from that major see a 1.7-percentage-point relative drop in their probability of securing employment in Texas within one year after graduation. Among graduates who do land jobs, their first-year earnings also fall by roughly 5% on a relative basis.

Does AI actually take away jobs? This debate has raged for two years with no simple answer. The overall U.S. unemployment rate has not spiked suddenly after the launch of ChatGPT, and there has been no massive wave of AI‑driven layoffs. Yet the picture looks far less optimistic when we focus on recent graduates.

AI’s impact on employment may not first appear in layoff rosters. When companies reduce hiring for junior roles and assign tasks once reserved for new hires to AI, the job market still shifts.

A study released by the Dallas Fed on September 22 delivers noteworthy findings.

The results show: for every 10‑percentage‑point increase in the share of automatable tasks associated with a given major, graduates from that major see a 1.7‑percentage‑point relative drop in their probability of securing employment in Texas within one year after graduation. Among graduates who do land jobs, their first‑year earnings also fall by roughly 5% on a relative basis.

Majors including Computer Science, Computer Engineering and language studies exhibit high exposure, while fields such as nursing, education and psychology show relatively low exposure.

This statistic should not be interpreted as “1.7% of graduates lose their jobs to AI.” It measures relative differences between majors with high and low AI exposure. The study sample is largely limited to graduates from Texas public universities and their employment records within Texas, so findings cannot be directly generalized to the entire United States.

Though the scope is narrow, graduates in high‑AI‑exposure majors have indeed begun to diverge from other majors in employment and earnings outcomes following ChatGPT’s arrival.

01 Why New Grads Feel the Chill First

When discussing technological displacement, people tend to track how many existing workers get laid off. But when it comes to AI, the hiring side may deserve more attention.

Many junior white‑collar roles consist of standardized work: organizing materials, drafting preliminary reports, processing routine data and writing basic code. Previously, firms hired new graduates to handle these tasks, and juniors built experience in the process. Now that AI can perform part of this work, companies naturally reassess how many junior employees their teams actually need.

An earlier Dallas Fed study published September 1 found AI automation exposure may reduce the volume of Texas job postings by approximately 1.8% in 2024 and 2.6% in 2025. Instead of mass layoffs, businesses can simply open fewer new roles, scale back campus recruiting, or reserve openings for experienced candidates.

This closely matches the reality many new grads are experiencing: jobs still exist on hiring platforms, but interviews grow harder to secure; job titles remain largely unchanged, while requirements quietly rise.

02 AI Makes Experience More Valuable

AI impacts workers unevenly depending on seniority.

A February analysis from the Dallas Fed notes that AI can more easily codify explicit knowledge — knowledge obtainable from textbooks, rules and documents. It struggles, however, to replicate years‑built judgment, client relationships and institutional experience.

Within the same job, AI can take over some standardized junior tasks and help senior employees deliver work faster. Companies still need people, yet they are more willing to pay for judgment, experience and accountability for outcomes.

Here lies the awkward paradox. Firms once used basic tasks to train new talent. As this category of work shrinks, young professionals lose opportunities to learn on the job. You cannot land a job without experience, yet you cannot gain experience without a job. This loop troubles new graduates even more than the question of whether AI will replace specific professions.

03 Sluggish Graduate Hiring Is Not All AI’s Fault

The Dallas Fed provides fresh evidence, yet weakening employment outcomes for new grads stem from multiple factors.

Related research from the New York Fed argues remote work has reduced employers’ willingness to train and mentor young employees, potentially explaining much of the rise in young graduate unemployment post‑pandemic. High interest rates and the tech industry digesting over‑hiring during the pandemic also weigh on new‑grad hiring.

Another study using national household survey data found no abnormal spike in U.S. new‑grad unemployment in summer 2026. This conclusion is not contradictory to the Dallas Fed report. The national aggregate unemployment rate can stay stable, while graduates in high‑AI‑exposure majors still face relatively weaker employment and earnings. The two studies examine different layers of the labor market.

Therefore, it is premature to claim “AI has destroyed U.S. campus recruiting.” A more accurate framing: U.S. hiring was already cooling, and AI has reshaped staffing demand for certain junior white‑collar roles. New entrants to the workforce are the first group to hit this headwind.

04 How Should International Students Prepare?

Listing ChatGPT or Claude in your resume skills section carries little weight now. Hiring managers want to know what problems you have solved using AI, how you catch errors in AI outputs, and whether you can translate outputs into real business outcomes.

Project and internship descriptions also need rewriting. Instead of listing how many analyses, scripts or pieces of writing you completed, clearly state the origin of the problem, why you selected a certain approach, and how you validated your results. AI can generate first drafts, yet it cannot explain why a decision is sound.

When targeting roles, avoid crowding into a small set of generic positions. The same technical skill set applied to healthcare, finance, manufacturing or energy tackles distinct problems. Industry domain knowledge, communication skills, and awareness of when AI tools fail are far harder to replicate than proficiency with one more AI product.

International students face an extra layer of practical constraints: work authorization, visa sponsorship, graduation timeline and recruiting cycles all must be considered together. Waiting until graduation to validate your career path leaves minimal room for adjustment. Internships, campus projects, job descriptions and networking serve to help you test whether a path is viable at an earlier stage.

Manteng: Do Not Wait Until Graduation to Realize You Need to Reset Your Career Direction

AI has changed not just headcount numbers, but also how companies screen junior talent. Risks vary across majors, internship backgrounds and immigration status. Manteng maps out hiring requirements tailored to each student’s graduation timeline, target roles and work authorization status. We identify missing evidence of competence in resumes, projects and interview prep, and refine target roles based on real market feedback.

We help students avoid making decisions based purely on “hot majors” or “trendy tracks,” and allocate limited preparation time toward more matched and actionable career paths.

This shift is still in its early phase; no one can precisely predict which jobs will disappear. Yet one truth is already clear for New Grads: the basic hands‑on tasks used for training are shrinking.

Going forward, a diploma alone will rarely guarantee your first job. Employers seek candidates who can leverage AI to get work done — and also know when not to trust AI.