The Trillion-Dollar Question

The secondary market already answered the question that the IPO will make official. Elite capital is not pricing Anthropic as a safety lab. It is pricing Anthropic as the structural winner of the enterprise AI segment — and it is doing so at a valuation that the primary markets have not yet caught up with.

In February 2026, Anthropic successfully closed a massive $30 billion primary funding round, cementing a post-money valuation of $380 billion and placing it among the most valuable private companies globally. However, these primary valuations rapidly proved insufficient to capture the intense market frenzy and fear of missing out surrounding the company. Due to the explosive, undeniable adoption metrics of Claude Code and the distinct lack of a public vehicle for retail and institutional exposure to pure-play AI software, unprecedented investor demand aggressively shifted to private secondary marketplaces such as Forge Global. On these secondary exchanges, existing Anthropic shareholders and early employees faced immense, daily pressure to offload equity, driving implied valuations to hover relentlessly near the monumental $1 trillion threshold. In extreme, documented instances, secondary shares were quoted and bid upon at valuations ranging from $1.05 trillion to $1.15 trillion. — Boris Cherney, Claude Code, Anthropic › The Architecture of Autonomy: Boris Cherny, Claude Code, and Anthropic's Trajectory Toward AGI and IPO › The Financial Horizon: Hyper-Scaling, Revenue, and the Imminent IPO › The Valuation Explosion and Secondary Markets

The gap between $380 billion and $1.1 trillion is not a rounding error. It is a signal — about the difference between what a primary round can price, with its information constraints and negotiated terms, and what the secondary market implies when sophisticated capital operates without those constraints. The secondary market is saying: the primary round underpriced it. And the secondary market, in this case, may be closer to right.

Editorial note: The corpus embedded for this book was assembled through early May 2026. Since that date, reporting indicates Anthropic has secured an additional fundraising round at a valuation approaching $900 billion — a figure that places the secondary market's $1 trillion+ implied price within months of becoming the primary-market consensus. The chapter's arguments are grounded in the corpus; the trajectory is noted here as post-corpus context.

The revenue story underneath the valuation is unusual. Frontier AI companies typically carry revenue that is structurally thin — dependent on one or two major relationships, exposed to model pricing compression, and not yet proven at enterprise scale. Anthropic's numbers do not fit this description.

Following a $30 billion Series G funding round led by institutional heavyweights such as GIC and Coatue, Anthropic achieved a post-money valuation of $380 billion. This valuation is anchored by a rapidly accelerating revenue engine; the company reported a $14 billion run-rate in February 2026, which subsequently surged to an estimated $19 billion by early March. This growth is heavily driven by deep enterprise penetration, with over 500 customers each spending in excess of $1 million annually. Furthermore, the company has developed highly lucrative secondary revenue streams; the Claude Code product line alone operates at an estimated $2.5 billion run-rate and commands roughly 4% of all public GitHub commits. — Anthropic Leak: Strategic Analysis › The Sovereign Singularity: A Multidimensional Analysis of Anthropic's $380 Billion "Accidental" Operating System Leak › The Trillion-Dollar Battlefield: Anthropic vs. OpenAI in the IPO Arms Race

$5 billion of run-rate growth in six weeks is not a normal SaaS trajectory. It is the revenue signature of a product that has crossed some threshold — not just adoption but embeddedness. The 500 customers spending more than $1 million annually are not experimenting with Claude. They have built workflows around it. Claude Code's 4% of public GitHub commits means the product is not a developer curiosity; it is inside production pipelines at a scale that would require active effort to remove. This is the switching-cost moat the commercial wedge chapter described as the real value of the Claude Code bet. The revenue is now proving the thesis.

The IPO structure forces a question the company has not had to answer in a private context: what does a publicly traded safety lab look like?

The Responsible Scaling Policy, as a governance mechanism, was designed for a specific kind of accountability — public commitments, revision under scrutiny, deployment gates that are legible and therefore costly to revise quietly. This is a reasonable governance architecture for a private company. It is a materially different architecture for a public company with quarterly earnings calls, institutional shareholders with fiduciary duties to maximize returns, and an S-1 that has to disclose the conditions under which Anthropic would voluntarily halt deployment of its most capable models.

The RSP's ASL-4 trigger — the condition under which Anthropic has committed to a deployment pause it has never had to execute — is not a risk that investment banks model easily. It is the kind of contingent liability that requires disclosure language, that will generate analyst questions, and that creates a structural tension between the company's governance commitments and its shareholders' reasonable expectation that management will not voluntarily halt the company's primary revenue-generating activity.

This is not a hypothetical. ASL-4 capability thresholds are approaching. Anthropic has said so publicly, in the RSP. The S-1 will have to describe what happens when they arrive.

The competitive framing around the IPO is its own pressure system.

The private market capitalization of these two frontier laboratories has reached sovereign levels, fundamentally altering the macroeconomic landscape of the technology sector. OpenAI closed a blockbuster $110 billion funding round in late February 2026 — bankrolled by Amazon, SoftBank, and NVIDIA — propelling its pre-money valuation to $730 billion and its post-money valuation to an astonishing $840 billion to $850 billion. To bridge this massive valuation gap and secure the necessary capital to fund the Broadcom TPU orders, Anthropic executives, guided by investment banks including Goldman Sachs, JPMorgan, and Morgan Stanley, have aggressively accelerated their IPO timeline. The company is reportedly targeting a public market debut as soon as October 2026, seeking to raise upward of $60 billion in what would be one of the largest offerings in financial history. — Anthropic Leak: Strategic Analysis › The Sovereign Singularity: A Multidimensional Analysis of Anthropic's $380 Billion "Accidental" Operating System Leak › The Trillion-Dollar Battlefield: Anthropic vs. OpenAI in the IPO Arms Race

The race to the public markets is not purely about capital. It is about the pricing of the category. Whoever prices first sets the multiple against which the second company is evaluated. If OpenAI lists at an $840 billion valuation with $13 billion in revenue, the implied multiple establishes the floor for Anthropic's own pricing — and Anthropic, with higher revenue per dollar raised and a more defensible enterprise position, will argue for a premium. The October 2026 target is aggressive in part because waiting is expensive: every month before the IPO is a month the category multiple is being established by a competitor.

The structural question the IPO does not resolve — the one the public benefit corporation structure was meant to address — is whether public shareholders are compatible with the safety commitments that define the company. The PBC structure gives Anthropic's board explicit authority to consider the public benefit mission in decisions that would otherwise maximize shareholder value. This is legally meaningful. It is not infinitely durable under the kind of institutional pressure that a trillion-dollar public company generates.

The founding argument — from chapter 2 — was that structure matters more than culture, and that structure-first governance would hold under pressures that culture-first governance could not. The RSP is the structure. The PBC is the structure. The October 2026 IPO is the first test of whether those structures hold when the shareholder register includes pension funds, index funds, and hedge funds whose fiduciary mandates are not oriented around frontier AI safety.

Editorial aside: The S-1 is the document that will settle this question at the level of legal language. What Anthropic writes about the RSP, about ASL-4 triggers, about the relationship between the PBC structure and shareholder rights, will be the most consequential statement of the founding wager's durability since the wager was made. The book will be out before the S-1 is filed. This is the chapter's honest limitation.

What the trillion-dollar question actually is: not whether Anthropic can achieve the valuation — the secondary market has already answered that — but whether the governance architecture that was designed for a 11-person safety lab can remain load-bearing at the institutional scale a public market listing demands. The founders built a structure meant to hold under pressure. The pressure is arriving in October.

The answer to the question will not be in the S-1. It will be in what Anthropic does the first time a major deployment decision is contested by shareholders who read the RSP differently than the founders wrote it.