AI/ML Jobs Hiring Report — June 2026

The market is still adding roles, and a lot of new names are showing up

Over the last 28 days the dataset gained a net 1,500 AI/ML roles — 3,558 opened against 2,058 closed. That's not a churn artifact; it's genuine expansion, and it puts the current total at 9,139 open roles across 2,792 companies pulled straight from their own Greenhouse, Lever and Ashby boards.

The number I keep coming back to is 431: that's how many companies posted their first AI/ML role in the last 28 days. Roughly one in six of the companies hiring right now wasn't on this board a month ago for this category. New entrants at that rate tell you the hiring isn't concentrated in the usual handful of labs — it's spreading into companies that previously had no measurable AI/ML headcount need, or at least weren't advertising one publicly.

If you want the live version of any of these figures, the running view sits at the AI/ML jobs hub, and the company-level churn is tracked on the movers page.

Who's scaling fastest right now

The fastest-hiring company by net new roles is a tie at the top: LILT and Mistral AI each added a net +17 over the window. The difference is in the base — LILT is sitting on 92 active roles versus Mistral's 71, so LILT is both larger and still growing, which is the more interesting signal of the two. Sustained hiring on an already-large board is harder to fake than a quick burst.

After that the pattern shifts toward companies building from a small base. Gen added +15 net and has 15 active — essentially everything they have open in this category opened recently. Sarvam (+14, 14 active), Nscale (+13, 13 active) and Clera (+13, 13 active) all show the same fingerprint: net new roughly equals total active, meaning these are standing-start ramps rather than steady-state replacement hiring.

The rest of the top ten leans toward established product companies topping up: Life360 (+13 net, 27 active), Coupang (+12, 30 active), Drata (+11, 15 active) and Roku (+11, 24 active). Worth saying plainly — these are net figures, so a company closing a stale req and opening two new ones still nets +1. The companies where active count and net new move together are the ones genuinely building, and most of this list qualifies.

This is an engineering market, full stop

If you came to AI/ML hoping the demand was for researchers or product people, the data is blunt: engineering accounts for 8,730 of the 9,139 roles. That's 95.5%. Data analytics is a distant second at 315, and everything after that — operations (7), design (6), product (5), sales (4) — is statistical noise. There are 11 roles the classifier flagged as junk, which is about what I'd expect, and I'd rather report it than quietly drop it.

The takeaway for anyone positioning themselves: AI/ML hiring in 2026 is the hiring of people who ship models and the systems around them. If your background is adjacent — research, analytics, product — the roles labelled as engineering are still where the volume is, so the question is whether your skills map onto an engineering job description, not whether a parallel non-engineering track exists at scale. It mostly doesn't, at least not on these boards.

Seniority skews experienced — but read the "unspecified" bucket carefully

The largest seniority group is "unspecified" at 4,485 roles, just under half the total. That's not a finding so much as a limitation: companies frequently don't put a clean level in the posting title or fields I can parse, and I won't guess one in. So treat the rest as the picture among roles that did disclose level.

Of those, senior dominates at 2,699, followed by principal at 1,186 and director at 380. Junior sits at just 286. Put differently: among levelled roles, senior-and-above outnumbers junior by well over ten to one. Executive (83) and VP (20) are small in absolute terms but their presence at all says some of this hiring is org-building, not just individual contributor backfill.

For early-career candidates this is the hard part of the report. The market is growing, but it's growing at the experienced end. 286 explicitly junior roles out of 9,139 is thin, and even allowing that some "unspecified" postings are open to junior applicants, the demand signal is unambiguously weighted toward people who've already shipped. If you're early-career, your realistic path is targeting the unspecified roles where requirements are negotiable, not hunting for a "junior ML engineer" title that mostly doesn't exist here.

What AI/ML roles actually pay

First, the honesty clause: only 15% of postings disclose pay — 274 of them in the band I'm reporting. So everything here describes the minority of companies willing to publish a number, and those skew toward US roles and jurisdictions with pay-transparency laws. It is not the whole market, and disclosed bands tend to run a little richer than the silent majority. With that caveat, the overall median band lands at $152,875–$216,061 USD/year.

Breaking it out by role, where the sample is large enough to mean anything:

• Machine Learning Engineer — $174,000–$253,500 (n=54). The clear top of the table, and the largest single role sample, which makes it the most reliable number here.
• Research Scientist — $150,000–$243,950 (n=26). Lower floor than ML Engineer but a wide band, reflecting how much the title spreads across levels.
• AI Engineer — $150,000–$203,500 (n=25). The most compressed of the three at the top end.

The thing I'd point out: the "Machine Learning Engineer" and "AI Engineer" titles are often used interchangeably in job ads, but the disclosed pay says they aren't priced the same. The ML Engineer band tops out roughly $50k higher. If you're choosing which postings to chase and both titles describe work you can do, the ML Engineer framing is worth more on paper. Sample sizes are small — 54 is decent, 25 is suggestive at best — so I wouldn't bet a salary negotiation on a $50k gap, but it's a real pattern in the disclosed data.

The fuller breakdown by role and level lives on the AI/ML salaries page, and it updates as new disclosed bands come in.

Remote is a third of the market, not the default

Remote roles make up 32.0% of postings. I flag this every edition because the expectation gap is real — a lot of candidates assume AI/ML is a remote-heavy field, and the boards say roughly two in three roles are tied to a location. If remote is a hard requirement for you, you're competing for a third of the listings, and given how engineering-heavy and senior-skewed this market is, that third is a meaningfully smaller pool than the headline 9,139 suggests.

I don't have a clean geographic breakdown to share this edition that I'd trust enough to publish, but the disclosed pay being denominated in USD and the company mix tell you most of this is US-centric, with a visible thread of non-US names — Mistral, Sarvam, Nscale, Coupang — among the fastest growers. The international entrants are part of why the new-company count is as high as it is.

Reading the churn: is this growth or noise?

Net +1,500 on 3,558 opened and 2,058 closed gives a closed-to-opened ratio of about 0.58. Roughly three roles close for every five that open. That's a healthy ratio — it means roles aren't just being recycled, and the board is genuinely deeper than it was a month ago. In a cooling market you'd see closures matching or exceeding openings; that's not what's happening here.

The combination I find most telling is the +1,500 net and the 431 first-time companies. Growth concentrated in existing players would suggest a few well-funded teams scaling. Growth spread across hundreds of new entrants suggests AI/ML hiring is broadening into the wider company base. The second is the more durable kind of demand, because it doesn't depend on a handful of teams keeping their foot on the gas.

One caution on my own data: a "net new" role reflects what's posted on company boards, not necessarily what gets filled. Companies open reqs they later quietly close, and a posting staying live isn't proof of an active search. The directional signal — more roles, more companies, openings outpacing closures — is solid. The exact 1,500 is a snapshot, and snapshots move.

What this means if you're job-hunting right now

Concretely, here's what the June data tells you to do:

• Lead with engineering credentials. 95.5% of roles are engineering. Whatever your background, your application needs to read as someone who can build and ship, because that's where essentially all the volume is.
• If you're senior, the market is in your favor. Senior and principal roles together (3,885 levelled) dwarf junior. Experienced engineers are what these companies are paying for, and the disclosed ML Engineer band — up to $253,500 — reflects it.
• If you're early-career, target the "unspecified" reqs. Explicit junior roles are scarce (286). Your better odds are in the large pool of postings that don't fix a level, where you can make the case on skills.
• Look hard at the fast movers, especially LILT and Mistral. A company adding +17 net on an already-large board is actively building teams, which usually means faster pipelines and more open headcount than a company posting one role. LILT with 92 active roles is the standout for sheer volume.
• Don't anchor on the headline salary. Only 15% of postings disclose pay, and those skew high and US-based. Use the $152,875–$216,061 band as a ceiling-ish reference for transparent employers, not the median you'll be offered everywhere.
• Decide early how hard your remote requirement is. At 32%, remote-only narrows your search by two-thirds in a market that's already tilted toward senior and engineering.

The short version: this is a growing market that rewards experienced engineers and pays well when it pays transparently, with a wave of new companies entering that should keep widening the opportunity set. Watch the movers month over month — the companies adding net roles fastest are the ones most likely to still be hiring when your application lands.