Is AI eating the entry level?

This index measures one thing: what share of the titles employers post are entry-level-titled. It does not prove displacement, because a shrinking share can mean fewer juniors hired, more seniors hired, or titles drifting. What it gives you is the measured series the argument has been missing, weekly, from the employers' own boards. Read a sustained fall in that share as the apprenticeship problem becoming visible.

Why does the entry level matter more than the total?

Because the entry level is how a profession reproduces. Senior practitioners are made, not hired: they start as juniors doing supervised work. If the bottom rung narrows for a decade, the shortage arrives a decade later, in the senior seats, and by then no policy can conjure the missing cohort.

A measured series, generated September 10, 2026

The Apprenticeship Index

How much of AI hiring is entry level, measured weekly from the titles employers actually post.

What share of AI jobs are entry level?

As of September 7, 2026, 3.9% of title-level AI governance postings at tracked AI employers are entry-level-titled (8 of 203). In physical AI the share is 6.5% (50 of 764). The series starts with the indexes themselves and is recomputed from every archived weekly snapshot, so the trend becomes visible as the weeks accumulate.

3.9%

entry-level-titled, AI governance

6.5%

entry-level-titled, physical AI

0 pts

AI governance since September 5, 2026

01 The series

Every snapshot, both indexes

Each row is one weekly crawl. The series is young; it is printed from its first point because the archive is the moat, and a year of these rows is a record nobody can reconstruct after the fact.

AI Governance Hiring Index3 snapshots

Snapshot	Entry titled	All title-level	Entry share
September 5, 2026	8	204	3.9%
September 6, 2026	8	203	3.9%
September 7, 2026	8	203	3.9%

Physical AI Hiring Index1 snapshot

Snapshot	Entry titled	All title-level	Entry share
September 10, 2026	50	764	6.5%

02 Why it matters

The apprenticeship problem, measured instead of argued

Every profession trains its next generation the same way: juniors do real work under supervision, and a decade later they are the seniors. The strongest argument that AI breaks this is not that it eliminates jobs in total; it is that it absorbs exactly the tasks juniors learn on, so the bottom rung narrows while the top looks fine. Forecasts of that erosion exist by the hundred. A measured series existed nowhere, so this page now publishes one.

The institutional answer is the same page's subject as our thesis: the uneven transformation. The personal answer is a credential that proves what a junior role used to prove: the readiness check and the first free module behind it.

03 Method

How the index is built

Every week both GAGE hiring indexes read the public job boards of their tracked employers, keep the postings whose title names the work, and classify each title's seniority with fixed rules (intern, junior, associate and coordinator titles count as entry). This index recomputes the entry-level share for every archived snapshot of both. The figures are never copied forward: each build recounts the archives.

What "entry" means

A title is entry-level-titled when it carries intern, internship, graduate, junior, associate or coordinator, under the same fixed seniority rules both hiring indexes use. Titles are a proxy: a company can hire a junior into a mid-titled seat and nobody sees it. The series measures what employers say in titles, consistently, which is what makes the movement meaningful.

Limits, stated plainly

Coverage is the tracked employers of the two hiring indexes, not the whole market, and the two indexes measure different slices (governance-titled roles at AI employers; technical roles at embodied-AI employers). A share can move because juniors vanish, because seniors multiply, or because an employer retitles. The series is young: no trend is claimed until the archive is long enough to carry one.

04 Data

Download and cite

One file, rebuilt from the archives on every run, free under CC BY 4.0: index.json.

Cite as: GAGE (Global Academy of Generative-AI Education), "Apprenticeship Index", generated 2026-09-10, https://www.gage.academy/careers/apprenticeship-index. Reuse under CC BY 4.0.

Questions

What share of AI jobs are entry level?: As of September 7, 2026, 3.9% of title-level AI governance postings at tracked AI employers are entry-level-titled (8 of 203). In physical AI the share is 6.5% (50 of 764). The series starts with the indexes themselves and is recomputed from every archived weekly snapshot, so the trend becomes visible as the weeks accumulate.
Is AI eating the entry level?: This index measures one thing: what share of the titles employers post are entry-level-titled. It does not prove displacement, because a shrinking share can mean fewer juniors hired, more seniors hired, or titles drifting. What it gives you is the measured series the argument has been missing, weekly, from the employers' own boards. Read a sustained fall in that share as the apprenticeship problem becoming visible.
Why does the entry level matter more than the total?: Because the entry level is how a profession reproduces. Senior practitioners are made, not hired: they start as juniors doing supervised work. If the bottom rung narrows for a decade, the shortage arrives a decade later, in the senior seats, and by then no policy can conjure the missing cohort.
How is the index built?: Every week both GAGE hiring indexes read the public job boards of their tracked employers, keep the postings whose title names the work, and classify each title's seniority with fixed rules (intern, junior, associate and coordinator titles count as entry). This index recomputes the entry-level share for every archived snapshot of both. The figures are never copied forward: each build recounts the archives.

The raw postings behind every figure: the AI Governance Hiring Index and the Physical AI Hiring Index, both measured weekly.