What AI Is Doing to Entry-Level Jobs
The bottom rung of the ladder is the part that is breaking.
The entry-level job was always a trade. The employer got cheap, willing help with the dull work. The junior got a seat close enough to the real work to learn how it was done. Nobody wrote it down anywhere, but both sides understood the deal.
The dull work is exactly what AI turned out to be good at.
A February 2026 Resume.org survey of 933 US business leaders found 21% of companies have already frozen entry-level hiring because of AI. Another 36% expect to stop hiring for those roles entirely by the end of the year. That is not a forecast about the 2030s. It is most of a market deciding something within about eighteen months.
It shows up first in the sectors with the most routine document work. A King's College London study in early 2026 found junior positions at AI-intensive firms down 5.8% year-on-year, while senior roles held roughly steady. The cut is happening at one specific height.
| Reported figure | |
|---|---|
| US firms that have frozen entry-level hiring due to AI | 21% |
| US firms expecting to stop hiring entry-level roles by end of 2026 | 36% |
| Year-on-year fall in junior roles at AI-intensive UK firms | 5.8% |
The obvious cost is the jobs that are not there this year. The less obvious one arrives in about a decade.
Senior judgment is not taught. It accumulates, from doing the basic version of the work badly, being corrected, and doing it again. Remove the basic version and you have not made anyone a better senior. You have removed the only route to becoming one. Firms cutting junior roles today are drawing down a stock of experienced people that they are no longer replacing.
Aneesh Raman, Chief Economic Opportunity Officer at LinkedIn, put it plainly in a 2025 New York Times op-ed:
Breaking first is the bottom rung of the career ladder.
Citing Brookings research, he noted that young adults who spend six months unemployed at 22 can expect to earn roughly $22,000 less over the following decade. A gap at the start does not close. It compounds.
Some employers have worked this out and gone the other way. KPMG has started giving new graduates tax strategy work that used to require three years of experience, on the basis that AI closes enough of the technical gap to make it feasible. The junior is not doing less demanding work than they used to. They are doing more, sooner, with a machine handling the parts that used to fill their first two years.
That is the shape of the job now: less execution, more oversight. Checking output rather than producing it, and being answerable for what goes out. It is the same movement as the broader shift toward hiring on demonstrated skills rather than credentials, arriving at the point in a career where it hurts most.
It is a harder first job than the one it replaced, and worth being honest about that. Reviewing work you have never had to do yourself is genuinely difficult, and "keep a human in the loop" is easy to say and hard to staff. But it is a real job, which is more than can be said for the vanishing alternative. For anyone starting out in 2026, the aim is not to be faster than the machine at the routine work. That contest is over. It is to be the person who can tell when the machine is wrong.
Sources
- Resume.org. 1 in 5 Companies Have Stopped Hiring Entry-Level Workers Because of AI. March 2026.
- King's College London. New Study Reveals Early Impact of AI on the UK Job Market. 2026.
- Fortune, on Aneesh Raman's New York Times op-ed "I'm a LinkedIn Executive. I See the Bottom Rung of the Career Ladder Breaking". May 2025.
- Brookings Institution. Youth Unemployment Is a Problem for Social Mobility.