--- title: "Anthropic files its S-1: enterprise AI goes public" date: 2026-06-04T08:00:00+02:00 language: en slug: 2026-06-04-anthropic-ipo-s1-bourse-ia-entreprise url: https://mathieuhaye.fr/blog/en/2026-06-04-anthropic-ipo-s1-bourse-ia-entreprise alternate: https://mathieuhaye.fr/blog/2026-06-04-anthropic-ipo-s1-bourse-ia-entreprise category: Business & Growth description: "Anthropic filed a confidential IPO draft on June 1, 2026, valued at $965bn. What its S-1 reveals about the real business of enterprise AI." --- # Anthropic files its S-1: enterprise AI goes public > Anthropic filed a confidential IPO draft on June 1, 2026, valued at $965bn. What its S-1 reveals about the real business of enterprise AI. - **Anthropic filed a confidential Form S-1 with the SEC on June 1, 2026**, the first formal step toward an initial public offering (IPO). - **The filing follows a $65bn Series H raise** that values Anthropic at $965bn, just under the symbolic trillion-dollar line. - **Anthropic's annualized revenue run-rate hit roughly $47bn in May 2026**, up from about $9bn at the end of 2025. - **In April 2026, Anthropic overtook OpenAI in paid enterprise adoption** for the first time: 34.4% versus 32.3%, per Ramp's AI Index. ## The facts Anthropic, the maker of the Claude model, confirmed on June 1, 2026 that it had submitted a draft Form S-1 registration statement to the U.S. Securities and Exchange Commission (SEC) for a proposed initial public offering (IPO). The filing is [confidential](https://www.anthropic.com/news/confidential-draft-s1-sec): it relies on Rule 135 of the Securities Act, sets neither the number of shares nor the price range, and states that the deal will depend on market conditions. The timing speaks volumes. The filing lands less than a week after a $65bn Series H raise that pushes Anthropic's valuation to $965bn, per [TechCrunch](https://techcrunch.com/2026/06/01/anthropic-files-to-go-public/). The lab's annualized revenue run-rate is around $47bn in May 2026, up from roughly $9bn a year earlier. Anthropic now joins OpenAI and SpaceX in the line of listings that could clear the trillion-dollar mark in 2026; SpaceX filed its public S-1 on May 20 and targets a Nasdaq listing on June 12 under the ticker SPCX. OpenAI, for its part, raised $122bn in March 2026 at an $852bn valuation and is preparing its own market debut. ## Why does an AI lab file to go public now? Because Anthropic is no longer a research lab: it has become a recurring-revenue software company, and an S-1 exists precisely to turn that traction into permanent capital. An IPO filing forces you to present predictable, audited accounts; you do not file an S-1 on a promise, you file it on a cash flow. Multiplying annualized revenue fivefold in twelve months, from $9bn to $47bn, is exactly the kind of curve public markets know how to price. The second reason is blunter: the cost of compute. Training and serving models at Anthropic's pace burns tens of billions of dollars a year in infrastructure. Private venture capital can cover part of that bill, but not forever. Public markets offer a deeper, more liquid funding source and let early investors cash out some of their stake. Filing now, confidentially, starts the SEC review without exposing the financials to rivals; Anthropic keeps control of when it flips to public. Finally, there is the market window. With SpaceX listing as early as June 12 and OpenAI on the same trajectory, 2026 is shaping up as the year of trillion-dollar listings. Nobody wants to show up after the window closes. ## The race to trillion-dollar listings Anthropic's filing confirms a deeper shift: AI labs are being financialized. In under three months, three players are chasing twelve-zero valuations, two of them pure AI. OpenAI sits at $852bn after its March 2026 raise; Anthropic at $965bn after the Series H. The question is no longer whether generative AI is an industry, but how many giant listings it can absorb in a single year. That race has a direct consequence for corporate customers. A publicly listed vendor is bound to quarterly transparency: margins, churn, customer concentration, compute dependence. For a CIO betting their back office on Claude or GPT, that is good news; they will finally be able to read their vendor's accounts rather than guess at its health. It is also a safeguard: a public vendor cannot quietly change pricing or terms without answering to shareholders. The flip side is the pressure of the quarterly result. Once listed, OpenAI and Anthropic will have to reconcile frontier research, expensive and risky, with the demand for rising margins. That tension will decide the final bill customers pay. The aggressive free tiers and unlimited credits used to win the market rarely survive a public shareholder base for long. ## What enterprise adoption actually reveals What makes Anthropic's IPO credible is not its valuation, but its grip on durable software budgets. In April 2026, Anthropic took the lead in paid enterprise adoption for the first time: 34.4% versus 32.3% for OpenAI, per Ramp's AI Index as reported by [VentureBeat](https://venturebeat.com/technology/anthropic-finally-beat-openai-in-business-ai-adoption-but-3-big-threats-could-erase-its-lead). Over one year, Anthropic quadrupled its enterprise penetration while OpenAI grew by just 0.3 points. The engine behind that swing has a name: Claude Code, Anthropic's agentic coding tool, now the fastest-growing product in its history. According to Menlo Ventures, Claude captures 42% to 54% of enterprise coding spend, against 21% for OpenAI. More importantly, Anthropic has moved beyond technical teams into finance, legal and research workflows. That is where the real value sits: a model that wires itself into a business workflow creates switching costs, and therefore recurring revenue. The lesson goes beyond Anthropic. The market no longer rewards flashy demos, but the measurable replacement of tasks inside hard-to-move processes. A generic assistant is one click away from replacement; an agent plugged into accounting, the CRM or the contract base, far less so. The same caution applies on the buyer side: Ramp's data mostly reflects US mid-market and growth-stage companies, not necessarily large enterprises or international markets. Anthropic's lead is real, but it is measured on a segment, not the whole planet. ## How this connects to my work This story of recurring revenue and models wired into real usage, I live it at a small scale. On my Bloomberg Dashboard, I run Claude Haiku 4.5 over a personal portfolio: the model reads the positions, cross-checks the news and produces a daily readout, for a few cents per run. What makes the tool useful is not the model's raw power; it is that it plugs into real data, inside a workflow I actually use. That is exactly the mechanism that tips enterprise adoption. A standalone model is a demo; a model connected to a CRM, a scoring base or an automation flow becomes a dependency, and therefore a recurring cost. When I set up an n8n automation for an SME or a job-offer scorer for a candidate, the value never comes from the model alone; it comes from wiring it into the right systems. Anthropic's S-1 tells the same story, in a trillion-dollar version: the AI that sells is the one that embeds itself into the daily work of a business function. ## Take-away An AI IPO is not triggered by a model, but by a revenue stream. Anthropic files its S-1 because Claude has become a recurring budget line in finance, legal and code. The real question for 2026: once AI vendors have to report every quarter, who pays the bill for the compute race? ## Frequently asked questions ### What is a confidential S-1 filing? An S-1 is the registration document filed with the SEC for a US initial public offering. The confidential version, allowed under Rule 135 of the Securities Act, starts the regulator's review without immediately publishing the financials. Anthropic submitted this draft on June 1, 2026. ### What is Anthropic's valuation in June 2026? Anthropic is valued at $965bn after a $65bn Series H round closed in late May 2026. Its annualized revenue run-rate is around $47bn, up from roughly $9bn a year earlier. ### Has Anthropic overtaken OpenAI in the enterprise? In April 2026, Anthropic took the lead in paid business adoption for the first time, at 34.4% versus 32.3% for OpenAI, per Ramp's AI Index. According to Menlo Ventures, Claude captures 42% to 54% of enterprise coding spend, against 21% for OpenAI. --- Source: [https://mathieuhaye.fr/blog/en/2026-06-04-anthropic-ipo-s1-bourse-ia-entreprise](https://mathieuhaye.fr/blog/en/2026-06-04-anthropic-ipo-s1-bourse-ia-entreprise) | Other language: [https://mathieuhaye.fr/blog/2026-06-04-anthropic-ipo-s1-bourse-ia-entreprise](https://mathieuhaye.fr/blog/2026-06-04-anthropic-ipo-s1-bourse-ia-entreprise)