1. Anthropic IPO: an early backer explains what the company sells
Matt Murphy, a partner at Menlo Ventures, said in a television interview on November 19, 2026, that Anthropic is ultimately in the business of selling intelligence and the capacity to innovate. He spoke one day after the Anthropic IPO priced at $83 a share and raised $81.3 billion, the largest initial public offering on record.
Menlo is among the earliest investors in Anthropic, the developer of the Claude family of artificial intelligence models, and it kept putting money into the company across multiple funding rounds. The firm's holding after the listing is estimated at roughly $25 billion, which makes Murphy one of the investors with the most at stake in how public markets judge the business.
His comments, though, dwelt less on the size of the payday than on what he sees as the forces behind it. Murphy concentrated on the nature of what Anthropic sells, the potential of its coding product Claude Code, and the company's long-running emphasis on trust and safety.
2. The numbers behind a record debut
The offering was priced on November 18, 2026. Anthropic sold 980 million shares at $83 each, bringing in $81.3 billion. That total exceeded the previous record of $75 billion, which SpaceX had set earlier in the same year, and it moved Anthropic to the top of the list of the largest stock market listings ever completed.
The shares trade on the Nasdaq under the ticker ANT. Demand on the first day of trading was strong at the start. The stock opened at $111.50, a jump of 34% over the offering price, meaning investors who were allocated shares in the offering saw an immediate paper gain of more than a third.
The initial enthusiasm did not hold at that level through the session. The stock cooled as the day went on and closed on November 19 at $97.17. That still left it 17% above the price at which the shares were sold, a solid first-day result by ordinary standards, but well short of where the first trade had printed.
The gap between the opening and closing prices is a reminder of how first-day trading in a heavily anticipated listing tends to work. The opening price reflects a rush of orders from investors who could not get shares in the offering itself. Once that early demand is met, the price settles toward a level where buyers and sellers are more evenly balanced.
For the company itself, the first-day swings change nothing about the money raised. Anthropic collected its proceeds at the $83 offering price. What happens to the shares afterward determines the gains or losses of the investors who hold them, including early backers such as Menlo whose stakes are now valued by the market every trading day instead of at occasional private funding rounds.
3. Who owns and controls Anthropic after the listing?
Control of Anthropic remains with its founders. They hold significant voting power through a special class of shares, so the people who bought stock in the offering have economic exposure to the company without a matching say over its decisions.
Structures of this kind are common among technology companies that go public while their founders are still running them. Ordinary shareholders take part in the gains and losses of the stock, while the direction of the company stays with a small group whose shares carry extra voting weight. For Anthropic's new investors, that means buying into the judgment of the existing leadership as much as into the business.
The shareholder register also features two of the largest technology companies in the world. Amazon and Google are both major shareholders in Anthropic, giving two Big Tech firms a substantial financial interest in the company's performance as a listed business.
Early venture investors make up another part of the picture, and Menlo Ventures is prominent among them. A stake estimated at around $25 billion after the listing represents the outcome of a bet the firm made and then kept adding to over several rounds.
4. What does Murphy mean by selling intelligence?
Murphy's argument is that Anthropic's product is not software in the traditional sense but intelligence, along with the ability to innovate that comes with it. The framing is a way of describing what customers are actually paying for when they use a model such as Claude.
A conventional software company sells a tool that performs a defined function: a database stores records, a spreadsheet calculates, a design program draws. An AI model developer sells something broader. The same underlying model can draft a contract, analyze a dataset, write a program or answer a customer's question, depending on what it is asked to do. On Murphy's account, the commodity being sold is the reasoning itself, and the individual applications are simply the forms it takes.
That view helps explain why a venture firm would keep investing in the same company round after round. If the product is a general capability and not a single application, the market it can address is not limited to one industry or one type of task. Murphy presented that breadth as the foundation of the business, more than any individual feature or contract.
The phrase also carries an implicit claim about durability. Tools can be replaced by better tools, and individual applications can be copied. A company that supplies the underlying capability on which many applications depend occupies a different position, because its customers are building their own products and processes on top of it. Murphy paired intelligence with the ability to innovate for that reason: in his description, what buyers obtain is not a finished answer to one problem but a means of producing new answers of their own.
5. Claude Code and the disruption argument
Murphy also pointed to Claude Code, Anthropic's coding product, and described its impact as potentially disruptive. He did not go into specifics about the scale of that disruption in the interview.
The product applies Anthropic's models to software development. Coding has been one of the areas where AI models have found the most direct commercial use, because the output can be tested: a program either runs and does what it should, or it does not. A tool that can take on meaningful parts of that work changes how software teams operate, and that is the kind of shift Murphy's remark alludes to.
For investors assessing Anthropic as a public company, the significance is that Claude Code represents a concrete product line, not only a general model offered to other businesses. Murphy's decision to single it out suggests he regards it as one of the main ways the company's underlying intelligence is turned into something customers adopt and pay for.
6. Why safety remains part of the pitch
The third theme of Murphy's remarks was trust and safety. He said Anthropic's emphasis on both remains critical as its models become more powerful.
The point is commercial as well as principled. Organizations that hand important work to an AI system need confidence that it will behave predictably and within limits. The more capable the models become, the more consequential their mistakes or misuse can be, and the more weight customers are likely to place on how carefully a developer manages those risks.
Anthropic has made that emphasis a defining part of its identity since its early days, and Murphy treated it as a continuing source of strength, not as a constraint on growth. In his telling, the approach becomes more valuable as the technology advances, because the stakes attached to trusting a model rise along with what the model can do.
It is notable that an investor discussing a record-setting listing chose to give safety equal billing with product and growth. Public market investors often focus on revenue and market share. Murphy's remarks suggest that, for at least one of the company's longest-standing backers, the way Anthropic manages the risks of its own technology belongs in the same conversation as the financial case.
7. What the debut means for investors
Anthropic's listing gives public market investors direct access for the first time to one of the leading developers of AI models. Until now, exposure to the company was available mainly to venture firms such as Menlo and to large corporate backers.
The pricing and the first trading session have produced a clear set of reference points. The offering price was $83. The first trade was at $111.50. The first close was $97.17. Each tells a different part of the story: what institutional buyers agreed to pay, what the most eager buyers were prepared to pay when trading opened, and where the market found its balance by the end of the day.
Murphy's interview offers the long-term holder's reading of those numbers. From the standpoint of a firm that invested early and repeatedly, the listing is a milestone but not the thesis. The thesis, as he set it out, rests on three things: that intelligence is the product, that Claude Code shows how that product can reshape an established field, and that the company's commitment to trust and safety grows in importance as its models do.
Whether public shareholders come to share that view will be tested over quarters, not days. They are buying into a company whose founders keep voting control, whose largest outside holders include two technology giants, and whose earliest backers, by Murphy's account, see the record-setting offering as confirmation of a bet on intelligence as a business.
