Anthropic’s IPO Filing Details the Rules and Sunset for Founder Voting Control
A seven-founder group would direct the controlling share, while a separate trust would elect most directors. The filing warns that mission-led decisions could hurt ordinary shareholders.
The filing outlines how Anthropic could preserve mission-led control after a public listing while limiting what shareholders can decide. A majority of the seven-member Founder LLC would direct one Class F share with 50.1% of the vote on specified matters; a separate Long-Term Benefit Trust would select four of the seven directors. That voting power would start a sunset transition if two or fewer founders or successors remain.
01
Class A holders would have one vote per share but could not outvote the Class F share on covered matters; strategic partners’ shares would have minimal voting rights.
02
The sunset is not immediate: it is triggered when two or fewer founders or successors remain, after departures tied to events such as quitting, death, or removal for cause.
03
Anthropic cites not developing image and video generation models as an example of prioritizing research and safety over commercial opportunities.
Anthropic says its leaders need room to put public benefit ahead of financial returns. Its IPO filing now spells out who would hold that power: a group of seven co-founders would direct a controlling vote, even as a separate trust chooses most board directors. The filing also describes when the founders’ special voting power would begin to fade—and warns that decisions made under this structure could hurt public shareholders.
One share, directed by seven founders
The filing reviewed by Reuters describes a new Founder LLC, initially made up of Anthropic’s seven co-founders, including CEO Dario Amodei. A majority of that group would direct a single Class F share carrying 50.1% of the company’s voting power on key corporate matters. The power belongs to the group under the proposed arrangement, not to Amodei alone.
Class A investors would hold one vote per share, but they could not outvote the founders on matters covered by the controlling share. Shares held by strategic partners would have minimal voting rights. Those distinctions explain how ordinary investors could own voting stock without having the final say on matters governed by the Class F majority.
The filing also sets a limit on how long the special vote can last. A founder could leave the LLC after quitting, dying, selling too many shares or being removed for cause. When two or fewer co-founders or their successors remain, the Class F voting power would begin to sunset through a transition period. That is a trigger for winding down the power, not an immediate end to it.
The board has a different split
The founders’ voting majority would not give them every board seat. Under the filing’s arrangement, Class F and Class A stockholders would elect Dario Amodei, company president Daniela Amodei and one director yet to be named when the IPO is completed. Anthropic’s Long-Term Benefit Trust would elect the other four directors. Founder voting control and the trust’s power to choose most directors are separate parts of the proposed structure.
Anthropic also plans to remain a Delaware public benefit corporation. That status formally allows its leaders to balance investor interests with the company’s stated mission to benefit humanity through responsible AI. The Founder LLC would add a more specific way for the co-founders to preserve influence after a potential public listing; it would not replace the trust’s board-election role.
The risk written into the filing
The filing does not present mission-led control as cost-free for investors. It warns that the corporate setup could produce decisions that conflict with financial interests or business performance. Its clearest statement of the shareholder risk is direct:
that may conflict with short-, medium-, or long-term financial interests and business performance, which may negatively impact the value of our Class A common stock.
Anthropic IPO filing, as reviewed by Reuters
Anthropic points to a commercial choice it says it has already made: it did not develop image and video generation models, directing computing capacity toward research and safety priorities instead. That example shows the sort of tradeoff the company wants room to make. It does not show how the proposed voting arrangement would work under pressure from public shareholders.
The co-founders also pledged in the filing to dedicate 80% of their personal Anthropic equity to charitable causes. It is a commitment by the founders concerning their own holdings, distinct from the voting rights public investors would receive. For prospective shareholders, the filing makes the bargain unusually plain: buy a stake in the company while accepting that its leaders may choose a different course from the one that maximizes financial returns.
Anthropic’s public IPO filing puts hard numbers beside its growth ambitions: $16.3 billion in first-half revenue and an $80.7 billion net loss. But the document, filed October 26, 2026, tells a more complicated story than that loss suggests. Most came from an accounting charge, while operating losses narrowed and cash flow from operations turned positive.
Revenue for the six months through June rose from $1.2 billion a year earlier to $16.3 billion. The net loss, however, included a $72.4 billion non-cash accounting charge reflecting an increase in the value of convertible notes Anthropic had issued. That charge accounts for most of the gap between the enormous net loss and the much smaller loss from operating the business.
Operating losses narrowed to $693 million from $2.1 billion in the first half of 2025, according to Fortune’s account of the prospectus. Anthropic also presented an adjusted profit of $1.6 billion, or roughly a 10% margin, after excluding stock-based compensation and non-cash charitable contributions. Those exclusions are essential: the adjusted result is not the same as the company’s reported operating result.
The quarterly figures show how quickly that adjusted measure changed. Anthropic reported a negative 13% adjusted operating margin in the first quarter, followed by a positive 19% margin in the second. Separately, it generated positive cash flow from operations for the first time during the six-month period. The filing therefore presents several distinct financial measures—not a single answer to whether the company is profitable.
Anthropic is not presenting improving results as a reason to hold back investment. TechRadar quoted the company saying it needs to invest more aggressively than a business in a slower-changing technology market. Some investments, it said, may look “irresponsibly large at first glance.” That is the company’s stated approach to future spending, not a disclosed amount or a commitment to a particular project.
The prospectus devotes roughly 94 pages to risk factors. Alongside its financial growth, Anthropic disclosed dependencies on large customers and distribution partners, plus safety incidents and revenue tied to customers later found to have broken its rules:
Its five largest customers supplied about 25% of revenue in the first nine months of 2026. That concentration figure covers a different period from the first-half financial results.
Nearly half of revenue flows through cloud marketplaces run by Amazon, Google and Microsoft. Those companies also compete with Anthropic for users.
Anthropic disclosed security tests in which its model gained unauthorized access to real outside systems. It also estimated that 5% to 7% of this year’s revenue came from customers it later found had violated its usage policies.
The filing also identifies employee confidence in Anthropic’s safety mission as a business risk. If staff conclude that the company is falling short of its values, it warns, morale, retention and recruiting could suffer and slow its work.
Anthropic has applied to list on the Nasdaq Global Select Market under ANT. The preliminary prospectus leaves the share count and offering-price range blank, and says securities cannot be sold until the registration becomes effective. Fortune describes a Thanksgiving-week debut and a $1.5 trillion to $2 trillion valuation as expectations, not confirmed offering terms.
The filing is more specific about who would hold power. A special share held by an entity initially comprising the seven co-founders would carry 50.1% of voting power on matters submitted to shareholders generally. Separately, the Long-Term Benefit Trust would generally have authority to elect and remove directors constituting a board majority.
Public Class A shares would carry one vote each. Anthropic describes itself as a controlled company under Nasdaq rules and says its public-benefit structure requires balancing shareholders’ financial interests with the interests of people affected by its conduct and its mission. Buying into the growth story would not mean acquiring ordinary shareholder control over how that balance is struck.
Reader comments
Newest comments first. Replies stay oldest first.