Anthropic filed a confidential Form S-1 on or around 1 June 2026, on public reporting, at a private valuation in the vicinity of $965 billion. OpenAI followed on or around 8 June 2026, at a reported reference valuation between $730 and $852 billion, with some commentary placing the eventual pricing above $1 trillion. Neither filing is fully public at the time of writing, and the numbers in this paragraph are the best available public reporting rather than the disclosures themselves. The prediction market Polymarket, on 30 June, priced the probability of an OpenAI listing by 31 December 2026 at roughly forty per cent. That number will move materially in the coming weeks as the confidential filings clear the review process and become effective.
The honest headline is that two of the largest listings in market history are forming at once, at valuations set inside a narrow private market and about to meet public scrutiny for the first time. The instinct in the commentary has been to reach for a verdict. The mature reading is that the S-1 is the instrument that permits the verdict, and the reader who wants a defensible personal view is better served by knowing what to read first inside the filing than by knowing what the commentators have decided in advance.
A confidential filing, in plain language
A confidential Form S-1 is a registration statement submitted to the Securities and Exchange Commission under the Jumpstart Our Business Startups Act, or the more recent equivalents, that permits an issuer to review the disclosures with the staff before making them public. The filing becomes public no later than fifteen days before the road show for the offering commences. Until then, the reported numbers are second-hand, sourced from people involved in the process, and should be treated as directionally correct rather than precise. The reader who acts on the reported numbers before the effective filing arrives is acting on a leak, not a disclosure. The distinction matters.
Once the filing is public, four sections of the S-1 are load-bearing for the artificial intelligence listings in particular, and they are the four this piece will walk through in order.
What to read first: gross margin at the segment level
The first load-bearing disclosure is the gross margin at the segment level. Enterprise software at scale has historically posted gross margins in the seventies. The artificial intelligence businesses, on the public commentary, operate materially lower because of inference costs, principally the cost of running the models on accelerator infrastructure to serve each request. The trajectory of the segment gross margin across the eight quarters preceding the filing is the single most important variable in the path to profitability. A gross margin that expands ten points over eight quarters supports a very different valuation than one that holds flat or compresses.
The reader will find the disclosure in the Management's Discussion and Analysis section of the filing, alongside a narrative explanation of the drivers. The two questions to hold in mind while reading are, first, whether the improvement is driven by pricing or by unit cost, and second, whether the improvement is expected to continue or is described as approaching a plateau. The two paths carry different implications for the valuation the reader is being asked to accept.
What to read third: the going-concern language and the use of proceeds
The third load-bearing disclosure is the going-concern language, together with the use-of-proceeds statement. A public-benefit corporation with disclosed losses in the tens of billions and a disclosed cash burn of a similar magnitude will present a going-concern narrative in the filing. The narrative will describe the cash on hand, the committed undrawn credit facilities, the anticipated proceeds of the offering, and the runway that the combination is expected to provide. The reader is looking for the number of quarters of runway the combination supports at the current burn rate, and for the assumptions under which the runway extends or compresses.
The use-of-proceeds statement sits alongside. A raise dominated by primary issuance, with proceeds allocated to compute capacity and research, signals an extension of the operating runway. A raise with a material secondary component, in which existing holders sell into the offering, signals a partial liquidity event for the insider base. The distinction is material for the alignment between the new public shareholder and the existing insider base, and the S-1 will state it plainly.
What to read fourth: the governance and the second principal
The fourth load-bearing disclosure is the governance section, together with the description of the public-benefit corporation charter. Both companies are public-benefit corporations. The practical meaning is that the directors are permitted, and in some readings required, to weigh stated mission commitments alongside the financial interests of shareholders. In a conventional corporation, a board that turns down a profitable contract on ethical grounds opens itself to litigation. In a public-benefit corporation with a stated safety mission, the same decision is more defensible.
For the ordinary investor, the implication is not that the structure is bad. The implication is that the structure introduces a second principal, the mission, alongside the shareholder. In the SpaceX listing, the second principal is the founder. In these listings, it is a stated public interest. In both cases, the minority shareholder owns less than the headline figure suggests, and the disclosure of the arrangement in the charter and in the governance section of the filing should be read in full before any order is placed. The reader is not required to disapprove of the arrangement. The reader is required to know what they are agreeing to.
The competitive picture the S-1 will not fully describe
The filings will describe the competitive landscape in the standardised form required by the SEC, principally by naming competitors and describing the basis of competition. The description will be accurate. It will also, by the nature of the disclosure, understate the pace at which the competitive picture is evolving.
Google DeepMind continues to ship frontier models at a cadence that, on the published benchmarks, places it among the leaders. Meta has invested at a scale that is, in absolute terms, comparable to either of the listed names, and is pursuing a different distribution strategy through the open-weights channel. The Chinese frontier laboratories associated with Alibaba, ByteDance and a small number of better-resourced national champions are publishing models that compress the technical lead the United States laboratories enjoyed in 2024 and 2025. The vertical artificial intelligence businesses in coding, in legal, in life sciences and in customer support are each generating meaningful enterprise revenue on top of one or more of the foundation models, and are capturing the workflow rather than the inference.
The bull case for the two listings rests on a continued narrow frontier and the pricing power that the frontier confers. The bear case rests on diffusion, on the commoditisation of the underlying model layer, and on the migration of value to the application layer. Neither case is settled by the listing. The reader who treats the two scenarios as live possibilities and sizes the position accordingly will be better positioned than the reader who treats the bull case as inevitable.
A prudent participation
LUMINAIRE does not publish price targets and does not recommend the purchase or sale of individual securities. Three observations, however, are worth carrying into the effective filings.
First, wait for the disclosure. The reported numbers are the best available public information, but they are not the filing. The reader who acts on the reported numbers before the effective S-1 arrives is acting on a leak. The wait is measured in weeks, not months.
Second, read the four sections above in order. Gross margin trajectory, related-party and concentration, going-concern and use of proceeds, governance and the second principal. Everything else in the S-1 is context for these four disclosures.
Third, size the position as a slice of the volatility budget of a diversified portfolio, not as the centre of it. Both listings are, by any reasonable base rate, going to trade with materially wider intraday and inter-day ranges than the median large-cap technology name in the first year after listing.
Where to do the work
Readers who want the full institutional treatment can read Cabier's companion release, Priced for a Monopoly, which sets out the disclosure taxonomy in the form the allocators are reading it. The path-to-profit model, which takes recurring revenue, growth rate, gross margin, operating expense growth and a target valuation and returns the implied break-even year, lives on CalculatorIQ. Both are linked at the foot of this article. The educational and illustrative purpose of all three pieces is the same, to support a reasoned personal assessment. Nothing here is investment advice. The desk used Anthropic in preparation for research synthesis. Editorial judgement, structure and conclusions are the desk's own.