AI Has Minted an $852 Billion Company, Yet It Still Isn't Listed.

How do wholesale investors get in?

OpenAI's March funding round valued the company at US$852 billion, and it still does not trade on any exchange, nor do Mistral, Perplexity or most of the companies building the AI industry around them. The one frontier lab to reach public markets, xAI, got there indirectly after merging into SpaceX, which listed in June 2026 in a record US$75 billion IPO. For the rest of the sector, capital is flowing through private rounds, secondary sales and special purpose vehicles, channels that until recently sat well outside the reach of individual investors and are now open to those who qualify as wholesale investors.

How can I invest in AI?

You can invest in AI through listed shares in companies with heavy AI exposure, through pre-IPO and secondary positions in private AI companies, or through venture capital funds that back AI startups at the seed and Series A stage.

Listed shares are the most liquid of these options, priced daily and easy to exit, and the choice has widened this year now that SpaceX, which owns xAI, trades alongside Nvidia, Microsoft and Alphabet. Much of the AI upside is already reflected in those prices, with Nvidia having become the first company to pass US$4 trillion in market value in July 2025. Venture funds sit at the opposite end, reaching companies at their earliest stage but typically locking capital away for seven to ten years, with returns that depend on one or two outliers carrying a portfolio in which most companies return little or nothing.

Pre-IPO and secondary investing occupies the ground between those two, and it is where most of the interest in private AI now sits. An investor buying at this stage is backing a company that already has substantial revenue and a priced funding round to benchmark against, with a listing or acquisition as the likely route to liquidity, and is taking that position before public market investors have the chance to buy in.

Do I need to be an accredited or wholesale investor to invest in private AI companies?

Yes. In Australia you need to qualify as a wholesale investor under the Corporations Act 2001, and in the United States the equivalent status is accredited investor. The two most common Australian tests are net assets of at least A$2.5 million or gross income of A$250,000 a year for each of the previous two financial years, both confirmed by a qualified accountant's certificate, and NonPublic's wholesale investor guide sets out how each test works.

The US definition under SEC Regulation D sets the bar at a net worth above US$1 million, excluding a primary residence, or annual income above US$200,000 (US$300,000 with a spouse). The SEC expanded the definition in 2020 to include holders of certain professional licences, which widened access at the margins.

These thresholds shape how the pre-IPO market is built, from the size of the minimum cheque to the disclosure documents that come with it. Any platform offering retail investors a stake in early-stage AI companies is either relying on an exemption or carrying regulatory risk that could ultimately affect the investor's claim on the underlying shares.

What AI companies can I invest in right now?

There are plenty of private AI companies worth investing in, but almost none of them are directly accessible to individual investors, because they raise capital from institutions and strategic partners and tightly restrict who can buy their shares. OpenAI is the largest, having closed a US$122 billion round at an US$852 billion valuation in March 2026, and Bloomberg has since reported its annualised revenue passing US$40 billion. Beyond the foundation model labs, the private AI market runs through inference chips and cloud infrastructure with Etched and Together AI, enterprise tooling with Fireworks AI and Cohere, and voice with ElevenLabs and Wispr Flow, alongside a fast-growing group of physical AI companies building models for robots and factories. PitchBook's Q2 2026 Venture Monitor found that AI companies took 86 per cent of US venture dollars in the first half of 2026. xAI, which sat alongside OpenAI at the top of the private market for most of the past two years, has since moved into listed markets through its merger with SpaceX, so exposure to it now comes through SpaceX shares.

The practical route in for individual wholesale investors is through platforms that source shares from existing holders, such as employees and early investors, or secure allocations in pre-IPO rounds. NonPublic is one of these, alongside secondary marketplaces like Forge Global and Hiive. Many of these platforms pool investor capital into a special purpose vehicle, or SPV, a single entity that holds shares in the target company on behalf of everyone in it, which brings individual minimums well below what a direct secondary purchase would require. What any platform can offer at a given time depends on which shareholders are willing to sell and on what terms, so the list of available companies changes from month to month and a particular name may simply not be on offer.

What does it actually cost to invest in a private AI company?

The cost depends on the route, with SPVs being the most common entry point for individual wholesale investors because their minimums are set deal by deal and typically sit far below what a direct purchase requires. Buying shares directly on a secondary marketplace usually means a much larger single transaction, and venture funds generally ask for substantial minimum commitments, with the strongest AI managers often closed to new capital altogether.

Fees make up the second part of the cost. SPVs commonly charge a setup or management fee and carry of around 20 per cent of profits above a hurdle, and over a holding period of five years or more those charges can move net returns a long way from the gross figure, so it pays to compare structures on a net basis before committing. NonPublic's pre-IPO guide walks through how these fees work in an Australian SPV.

Secondary marketplaces build their costs into the spread on each trade, and because their deal flow is aimed largely at institutional and US accredited buyers, availability for any given company can be patchy.

What returns have private AI investors made?

The early backers of today's leading AI companies are sitting on some of the largest paper gains in venture history. Khosla Ventures, which put US$50 million into OpenAI in 2019, is regularly cited for having made one of the defining venture bets of the decade, and the investors who followed it into OpenAI's early rounds have seen their positions marked up many times over as the company's valuation climbed toward US$1 trillion.

Paper gains and realised returns are different things, and almost none of these positions have yet been through a liquidity event. OpenAI is widely reported to be preparing for a public listing, but until that listing happens and lock-up periods expire, the gains remain marks on a fund's books. A company can carry an extraordinary private valuation and still deliver a disappointing outcome if its IPO prices below expectations or the path to liquidity takes years longer than investors planned for, which is one of the central private market risks to weigh before committing.

Secondary pricing is one of the better real-time signals available. When shares trade below the last primary round, buyers are expressing doubt about either the valuation or the timing of an exit, and when they trade above it, demand is outrunning the stock on offer. Different AI companies have sat on both sides of that line over the past two years.

How do I value a private AI company?

Private AI companies are mostly valued on revenue multiples benchmarked against listed peers, since the largest of them are still unprofitable and growing too quickly for a discounted cash flow model to say much of use. OpenAI's US$40 billion revenue run rate against its US$852 billion valuation implies a multiple of roughly 21 times revenue, while mature, profitable software businesses such as Salesforce trade at single-digit multiples. Whether that gap reflects a fair premium for extraordinary growth or an excess of investor enthusiasm is the central question in AI valuation, and NonPublic's valuation guide explains the mechanics in more depth.

Strategic investors carry a lot of weight in these valuations. When Nvidia and Cisco joined xAI's round in January 2026, before xAI merged into SpaceX, or when Microsoft deepens its commitment to OpenAI, the backer is a company that knows the business from the inside as a partner, supplier or customer, and its participation tells outside investors a good deal about how the industry itself rates the company.

What should I look for before investing in an AI company?

Start with the business rather than the technology. Few investors can judge the quality of a model architecture, but anyone with access to the right information can assess how fast revenue is growing, how concentrated it is among a small number of customers, how quickly the company is burning cash and whether the founding team is still in place.

AI also has a particular moat problem, because models improve so quickly that an advantage built on one generation of capability can disappear within a year. Open-weight models from Meta and from Chinese labs such as DeepSeek have closed much of the gap with frontier models on several benchmarks, and the companies that have held their position tend to be those with proprietary data or deep enterprise integrations that customers find costly to unwind.

The shape of the revenue deserves close attention as well. A company with a hundred enterprise clients on large contracts carries a very different risk profile from one relying on millions of low-priced consumer subscriptions, and because inference is expensive to run, gross margins in AI can look nothing like those of traditional software. Businesses that serve AI at scale while keeping margins healthy build an advantage that compounds as they grow.

Team stability is harder to judge from outside and can change a company's prospects almost overnight. When Nvidia licensed Groq's technology in December 2025, in a deal reported at around US$20 billion, and hired founder Jonathan Ross along with other senior leaders, the company that remained was a very different business from the one its investors had originally backed, which is why knowing who built the product, and whether they are still there, belongs at the centre of any due diligence.

Is now a good time to invest in AI companies?

Private AI is expensive by any historical measure, with the largest labs valued on the assumption that they will hold dominant positions for decades and far more capital chasing allocations than there are shares to go around. In a market priced this richly, the entry price and the structure of the investment carry more weight than usual, because the gap between a well-bought position and a poorly bought one widens as valuations climb.

Concentration adds a further consideration. The same PitchBook data that shows AI taking most US venture dollars also shows capital pooling in the largest rounds, and while that reflects strong conviction among institutional investors, it also means most AI startups outside that group are more likely to be acquired or wound down than to reach an independent listing.

For a wholesale investor, the more defensible approach is to back specific companies with an identifiable competitive advantage, at a price that can be justified from the fundamentals, through a structure that gives clear legal title to the underlying shares, since each of those conditions that is missing adds risk the headline valuation will not show.

Look at the opportunities currently available on NonPublic.

NonPublic Pty Ltd (ABN 49 607 216 928) holds Australian Financial Services Licence #482668. Investments are available to wholesale and sophisticated investors as defined under the Corporations Act 2001. This content is general in nature and does not constitute financial product advice. It does not take into account your objectives, financial situation, or needs. Investing in private markets involves significant risk, including the potential loss of your entire investment. Past performance is not a reliable indicator of future results. You should obtain independent financial advice before making any investment decision.