Humain's Saudi AI IPO Is Real — There's No Stock to Buy Yet, So Here's the Yardstick to Judge It By

Saudi Arabia's sovereign-backed AI company Humain has quietly started doing the paperwork that precedes going public. Chief executive Tareq Amin is recruiting an IPO-preparation team, hunting for people with top-tier management-consulting credentials, financial and strategic depth, and — explicitly — previous experience taking companies public. For a retail investor whose feed lights up with "AI IPO," the obvious question is whether this is a chance to get in early on a national AI champion. Working through what's actually disclosed, the honest answer is: not yet. And understanding why is the useful part.

Humain is not a garage startup. Saudi Arabia's Public Investment Fund launched it in May 2025 as a full-stack national AI player spanning data centers, cloud infrastructure, models, and sector applications. The capital deployment before any public spreadsheet is staggering: roughly 600,000 planned Nvidia GPUs, $23 billion in technology agreements, a $5 billion-plus partnership with AWS, $3 billion plowed into Elon Musk's xAI (a stake that converted into SpaceX shares when the two merged), and a fresh $2.5 billion fund to build data centers, with ambitions for a venture fund that could top $10 billion.

Two things about that picture deserve attention.

First, what the market has not been told: essentially anything you would score. No revenue, no valuation, no share count, no customer concentration. In my process a number means nothing in isolation — valuation, growth, and profitability only become actionable next to sector peers. A company you cannot yet rank against a comparison set is one you cannot rate, no matter how loud the story is. That disclosure gap is the whole ballgame, and it's why the headline, taken as a trading signal, overreads the event.

Second, the news itself is early-course. This hiring push is preparation, not a debut. Earlier reporting described a dual listing on Saudi Arabia's Tadawul and the U.S. Nasdaq within three to four years — in other words, "IPO preparations" here means laying groundwork a couple of years out, not a next-quarter offering.

What a retail investor can do today is build the yardstick, because AI infrastructure is not a mystery — there are listed analogs already pricing the same wager. The nearest public relative to Humain's core business, selling rented Nvidia compute, is CoreWeave: it trades around 6.5 times trailing sales and roughly 10 times enterprise value to sales, and it's GAAP-unprofitable, with revenue up about 115% year over year against a free-cash-flow margin near negative 170% as it sinks cash into capacity. Nebius, another pure AI-infrastructure name, is far more extreme, near 45 times sales. That is the template for this corner of the market in 2026: the price is set on growth and projected demand for compute, not on current profit. These are growth-and-expectation stocks, and a Saudi champion listing into that same lane would be held to the same standard.

So when the prospectus finally lands, hold Humain to the questions you'd ask any of these names, because the business model shares the same stress point. First, revenue growth against the buildout: are those GPUs earning their keep, and at what utilization? Second, margin trajectory — rented compute is a capacity-hungry, thin-margin business at scale, and flattering EBITDA margins can be an artifact of how depreciation is handled. Third, the gap between the story and free cash flow, which is where AI-infrastructure hype usually dies. Fourth, a governance question unique to Humain: with the PIF majority-owned and state oil giant Aramco taking a minority stake, what does the free float actually own, and on which exchange? A sovereign majority with a thin minority float is a structure question, not a business one.

None of that gives you something to buy. There is no ticker, retail investors cannot own Humain directly, and the only public ways in — Aramco, Saudi Telecom, the contractor Al Moammar — are unrelated businesses, not pure plays on the AI champion. In portfolio terms that makes Humain a watch-list item for the growth sleeve, and only a conditional one: it becomes investable the moment financials are disclosed and it can actually be scored. That conditional is not a dodge; it's the discipline. The AI story will keep making noise, and the process job is to keep a company with no disclosed numbers in its proper place — a candidate to evaluate once the numbers exist, not something to chase on the strength of a hiring announcement.