Confidential S-1 Filings: OpenAI follows Anthropic's lead, and the SEC will uncover what private valuations were hiding — ActuIA

February 13, 2015 | Source: ActuIA
Create a watch on this topic Receive upcoming developments if any, with an AI-powered watch set up from this article.
Create my free watch

On June 9th, OpenAI announced that it had submitted a confidential S-1 filing to the SEC, the American securities regulator - eight days after Anthropic, which had officially filed its own document on June 1st. OpenAI's announcement cultivates carelessness: "We recently submitted a confidential S-1. We expect it to leak, so we are announcing it ourselves. We have not decided on the timeline." Behind the tone, a shift: the two labs that defined the AI decade are almost simultaneously embarking on the path to public listing.

What a confidential S-1 is - and is not

The confidential filing allows a company to have its prospectus reviewed by the SEC in secrecy, and to iterate on the regulator's comments without exposing its financials or risk factors. It does not equate to a decision to go public: the timeline remains open, and OpenAI as well as Anthropic refrain from setting one. However, the rule is clear on one point: before any offering to investors, the document must be made public, including financial statements.

What the documents will have to reveal

This is the real issue at stake. For years, the economics of AI labs have been read exclusively through unverifiable numbers: private funding rounds, annual revenues reported without audit, and negotiated valuations on a handshake basis. A public S-1 filing will end this blind spot. It will include audited financial statements—revenues, losses, cash flow—the risk factors, including the computing commitments made to cloud providers, the capital structure, and governance rights, as well as executive compensation.

For companies whose capital needs are driven by the cost of training and inference, the section on off-balance-sheet commitments—how many billions of compute hours contracted, for how long, from whom—will likely be the most scrutinized part of the document.

Why now, and why with an eight-day gap?

The near-simultaneous filings are not a coincidence of administrative timing. Both labs are raising funds from the same private markets, from investors with limited capacity, for computing needs in the tens of billions. The stock exchange is the only capital reservoir of an order of magnitude larger. And the first to go public will set the valuation benchmark for the other—neither can let the other write this benchmark alone.

What to watch for

Three things, in order. The publication of the documents themselves, which will occur weeks before any listing and constitute the first audited look at the real economics of the labs. The treatment of the atypical governance structures of the two companies, which institutional investors will need to evaluate. And the reaction of computing suppliers and strategic partners, whose agreements will, for the first time, be listed as opposing risk factors.

Add ActuIA to my preferred Google sources

Our articles will prioritize appearing in Google's "Top Stories."

Also read
  • GPT-6 Astra : ce qui a changé chez OpenAI en quinze jours New model GPT-6 Astra: What Changed at OpenAI in Fifteen Days September 10, 2026
  • Une équipe de trois chercheurs a compromis le dépôt interne d'OpenAI avec Claude Opus 5, pour moins de 3 000 dollars de tokens Malicious Use (deepfake, fraud...) A team of three researchers compromised OpenAI's internal deposit using Claude Opus 5 for less than $3,000 in tokens September 18, 2026
  • OpenAI filigranera ChatGPT dans l'UE mais laisse l'API en option New Feature OpenAI will watermark ChatGPT in the EU but leaves the API as an option October 6, 2026
  • Agents OpenAI sur un wiki : la Commission européenne confirme avoir reçu un rapport Malicious Use (deepfake, fraud...) OpenAI agents on a wiki: the European Commission confirms it has received a report September 8, 2026
In the Watch
  • 4 nov. According to Reuters, OpenAI employees have raised concerns about internal security for years: accessible code, unaudited tools, documents on GitHub and Slack
  • 3 nov. OpenAI raises its Security Bug Bounty rewards and prepares an internal security sprint with Cathedral Systems
  • Nov. 3 S-1 amended by Anthropic : 17 billion dollars in revenue for the third quarter, 895.4 billion compute commitments, alert on recursive self-improvement
All night →
ChatGPTopen source
ST
Stephane Nachez

ActuIA Editing - news, data and analysis on artificial intelligence for decision-makers.

@TAGn@