Is my AI-exposed major really hurting my first paycheck?
Yes, the early evidence points that way, though the effect is measurable rather than catastrophic. A 2026 working paper from the U.S. Census Bureau's Center for Economic Studies (working paper CES-WP-26-56, 'Graduating into Disruption: Labor Market Outcomes for AI-Exposed College Majors,' published September 10, 2026) found that graduates in the most AI-exposed decile of college majors saw their likelihood of initial employment fall by five percentage points, and their full-quarter initial earnings decline by roughly 13%, relative to less-exposed majors after ChatGPT's introduction in late 2022.
That 13% drop is significant. The study's authors, as reported by Staffing Industry Analysts on September 14, 2026, characterized the earnings decline as comparable in magnitude to the losses associated with graduating into a large recession. Part of what drives the number is not that graduates cannot find work at all, but that some end up moving into lower-paying sectors such as retail and restaurants. In other words, the risk you feel is real: the pull toward a stopgap job that pays less and drifts away from your training.
Understanding the mechanism matters because it tells you where to push. The problem is not that your degree is worthless. It is that the on-ramp into your field has narrowed, and when the on-ramp narrows, people take the nearest exit. Your job is to keep looking for the on-ramp a little longer and a little smarter than the graduate who gives up after the tenth rejection.
Why is it so hard to land an entry-level job right now?
Because the definition of "entry-level" has quietly shifted, and the broader market for new graduates has softened. Two forces are working against you at once, and neither is entirely about AI.
The first is the erosion of true entry-level roles. According to Indeed's chief economist, speaking at the Fortune Workplace Innovation Summit in 2026 and reported via IBISWorld on June 19, 2026, employers are increasingly moving away from hiring entry-level candidates, with a growing share of job postings that carry the entry-level label nonetheless asking for three to five years of experience. That contradiction is maddening for a new graduate, but it is useful information. It tells you that raw willingness is not enough; you need to close the perceived experience gap before an employer will look past the label.
The second force is a thinner job market overall for young graduates. A Federal Reserve Bank of St. Louis analysis, reported by The Epoch Times on September 15, 2026, put the unemployment rate for recent college graduates ages 22 to 27 at 5.6%, notably above the overall rate of roughly 4.1 to 4.3%. That same analysis attributed the biggest contributor to a decline in job openings, with AI displacement a smaller factor. This is worth sitting with: the St. Louis Fed's reading suggests your field feels frozen partly because hiring in general has cooled, not solely because a model is doing your future job. A cooling market rewards patience and preparation; it does not necessarily reward panic.
How do I get a first job in my field instead of defaulting to retail?
Start by building visible proof of skill, then aim your applications at the roles most likely to still hire humans who can direct the tools.
Proof of skill is your fastest way to answer the three-to-five-years problem that the Indeed chief economist described. You cannot manufacture years of employment, but you can manufacture evidence. Ship a small portfolio of finished work: a deployed application, a documented data pipeline, an information systems audit of a local nonprofit, a public write-up of how you solved a real problem. Finished and public beats polished and hidden. When an employer's posting asks for experience they know a graduate cannot have, a demonstrable project lets them justify the exception.
Next, use the exposure research to your advantage rather than your discouragement. The Census Bureau study measured exposure by major, but exposure within a field is uneven. The tasks most at risk are the ones a model can complete without much context: boilerplate code, routine reporting, first-draft documentation. The tasks that remain human-directed are the ones requiring judgment, stakeholder translation, systems thinking, and accountability for outcomes. Position yourself toward that second category. In interviews and materials, talk less about writing code and more about diagnosing problems, choosing tradeoffs, and verifying that automated output is actually correct. That framing signals you are someone who supervises the tools, not someone competing with them.
Widen your target list to adjacent and less-exposed roles that still use your degree. A software or information systems graduate can pursue implementation, technical support engineering, solutions roles, data operations, QA, and internal tooling positions inside less-exposed industries such as healthcare, logistics, energy, or government. These roles keep you in your field and on a salary track, and they often have less competition than the small pool of glamorous first jobs everyone chases. Since the St. Louis Fed identified fewer openings as the main drag, casting a wider net across sectors directly counters the constraint you are facing.
Finally, treat the stopgap decision deliberately. If you must take a bridge job to pay rent, protect your trajectory by keeping it part-time or short-term, and continue applying and building in your field on the side. The danger the Census data hints at is not taking one retail shift; it is drifting into a lower-paying sector and staying there because the search stalled. Set a review date, keep your portfolio growing, and keep at least a few field-relevant applications going out every week. The graduates who recover fastest from a soft market are usually the ones who never fully stopped aiming at the field they trained for.