Anthropic Officially Launches IPO Roadshow: 825 Million Shares Priced at $70–$78, PBC Structure Poses Governance Risks

On November 10, 2026, Anthropic, PBC officially announced the launch of the roadshow for its initial public offering of Class A common stock. The offering totals 825 million shares, with a price range of $70 to $78 per share. The company plans to list on the Nasdaq Global Select Market under the ticker "ANT," intends to set the offering price on November 18, and expects trading to begin on November 19. At the top of the range, the IPO could raise approximately $64.35 billion, with a fully diluted valuation of up to $1.4 trillion.

From $96.5 Billion to $1.4 Trillion: A Rapid Valuation Leap

According to multiple reports, Anthropic completed a $65 billion Series H funding round in May 2026 at a valuation of approximately $96.5 billion, already surpassing OpenAI's private-market valuation of about $85.2 billion at the same time and making it the world's most valuable private AI company. Just six months later, the implied valuation in its IPO roadshow has jumped to the $1.4 trillion range.

Fortune reported that Anthropic had previously signaled a $2 trillion valuation target. The drop from $2 trillion to $1.4 trillion—a cut of nearly 30%—reflects the capital market's systemic caution toward high AI valuations. Several AI-linked stocks have corrected over the same period, and the roadshow is taking place against the backdrop of an "AI stock selloff."

Supporting this valuation are rapidly rising revenue figures. According to a summary of SEC filings disclosed by StockTitan, Anthropic's quarterly revenue grew by as much as 1,127% year over year, and its annualized revenue has reached approximately $47 billion. By comparison, OpenAI's annualized revenue over the same period is about $24 billion. Anthropic has captured about 55% of the enterprise large language model API market, according to estimates by Menlo Ventures, largely thanks to the scaled deployment of the Claude series in enterprise coding assistants and API integrations.

Three Unconventional Arrangements in the Offering Structure

Beyond the specific terms in the roadshow announcement, three arrangements warrant separate scrutiny.

First, strategic investors have locked in ahead of time. CNBC reported that seven strategic partners, including Google and Nvidia, have agreed to buy a total of about $26 billion in non-voting shares at the IPO price. This means that before public trading begins, institutional strategic demand has already absorbed more than 40% of the total offering, and the float actually available to retail investors is far smaller than the nominal offering size.

Second, the retail allocation is 20%, above convention. CryptoBriefing reported that Anthropic plans to allocate about 20% of IPO shares to retail investors, a relatively high proportion for a tech IPO and far above most high-profile technology listings. The decision sends a dual signal: it deliberately gives the public a sense of participation, while also proactively opening the retail channel to ensure secondary-market liquidity after strategic buyers have taken up a large share of the institutional allocation.

Third, the underwriters hold a 30-day over-allotment option. Goldman Sachs and Morgan Stanley may purchase up to an additional 123.75 million shares within 30 days after listing. If fully exercised, the total IPO size would increase further. This is a standard "greenshoe" mechanism and, amid current market volatility, also serves as a buffer to stabilize the share price.

PBC Structure: Listed, but Governance Control Does Not Belong to the Public Market

Anthropic is a public benefit corporation (PBC), a special corporate form under Delaware law that requires the board to balance three interests when making decisions: shareholders' financial returns, the interests of all parties affected by the company's actions, and the public benefit mission explicitly written into the company's charter. In Anthropic's charter, that mission is stated as "the responsible development and maintenance of advanced AI for the long-term benefit of humanity."

Public-market investors who buy Class A shares do not have the right to independently elect board members. A majority of board seats are selected by the Long-Term Benefit Trust (LTBT), which consists of five domain experts. These experts hold no company equity and are not affected by share-price fluctuations. At the same time, a single Class F share held by the seven co-founders carries 50.1% voting power on most matters.

This means that even if institutions and retail investors together hold more than 80% of the outstanding shares, their voting influence on major matters such as corporate strategy and management appointments remains lower than the one vote held by the seven founders through a single special share. Such dual or even triple-class control structures are not uncommon among listed technology companies (Google and Meta both have super-voting rights designs), but Anthropic's version is more extreme: it places the ultimate constraining authority in a trust that is completely detached from market incentives, rather than in the founders themselves.

In an in-depth report before the IPO, CNBC noted that this structure "creates unusual risks for investors." A Columbia Law School blog also published an analysis arguing that after listing, the PBC structure will face dual compliance pressure from shareholders and courts—when the AI safety mission conflicts with quarterly profit targets, the legal texts currently provide no clear answer.

A Mirror Comparison with OpenAI

Anthropic's listing must inevitably be understood in comparison with OpenAI. The two companies split from the same group of founders but have taken sharply different paths in technology, business model, and capital structure.

In terms of revenue, according to multiple data sources, Anthropic's annualized revenue is about $47 billion, compared with about $24 billion for OpenAI, giving Anthropic the lead. In terms of valuation, the $1.4 trillion upper bound implied by Anthropic's IPO is far above the $85.2 billion private-market valuation corresponding to OpenAI's March 2026 funding round. Both companies are pursuing trillion-level capital returns in the same time window, but Anthropic has chosen to take this step through the public market, while OpenAI has yet to start an IPO process.

There is a signal in the timing: Anthropic has chosen to list first—while its revenue is ahead and the market has not yet fully digested the risks associated with AI concepts—to seize the narrative window. Once OpenAI follows, the two companies will compete directly for institutional capital in the public market, and the first lister will have a first-mover benchmark pricing advantage.

Independent Judgment

From a capital-markets perspective, Anthropic's listing is a carefully designed piece of timing: with revenue growing rapidly, private valuation already at a historical peak, and strategic shareholders locking in $26 billion of demand, the systemic risk of the IPO has been substantially reduced. The backing of Goldman Sachs and Morgan Stanley, along with the relatively high retail allocation, are pre-built safeguards for secondary-market liquidity.

But the real unknown is not the opening price; it is the first stress test of the PBC structure after listing. When AI regulation tightens, when a Claude deployment case sparks public controversy, and when quarterly profit pressure comes into direct conflict with safety research spending, how will the company choose between the charter mission tied to the Long-Term Benefit Trust and Wall Street analysts' EPS expectations?