Anthropic is going to conduct an IPO, but the capital can't wait any longer.

Early this morning Beijing time, Anthropic officially announced that it had confidentially submitted an S-1 prospectus to the U.S. Securities and Exchange Commission (SEC).

Just 10 days ago, SpaceX also submitted a prospectus, and then OpenAI released news saying that they also plan to conduct an IPO.

Well, all of a sudden, they are all going for an IPO.

But many people haven't noticed one thing. Before these companies went public, there had already been frequent cases of employees selling old shares, institutions reselling quotas, and SPVs flying everywhere - the trading activity was higher than that of many listed companies.

They are more like listed companies than actual listed companies.

It's not just overseas. You may have also seen in some groups messages asking to buy old shares of the domestic large model company "Dark Side of the Moon", which is still in the game but the only one not listed, even though there is no official channel. ByteDance, the "cosmic factory" that has been recognized by the market for many years but has never gone public, has always been the dream target for people to buy old shares for many years.

More and more institutions are starting to layout in both the primary and secondary markets; more and more startups have gained market attention and pricing power close to that of listed companies before going public.

There's no time to wait. Good projects are rising too fast, and time waits for no one. The money in the primary market should be invested not in the future, but in the present which is the future.

VCs are acting like PEs, wanting to exit; the primary market is acting like the secondary market, chasing consensus; the secondary market is acting like the primary market, with prices being realized before value.

Amid the AI boom, chaos in the process of integration is occurring.

01 The Primary Market Resembles the Secondary Market

For a long time in the past, there was a clear division of labor in the capital market: the primary market represented by VCs was in the business of "belief" - placing bets on companies before they were proven; the secondary market was in the business of "verification" - deciding the value of a company after it went public, disclosed data, and realized growth.

So the former bets on five years later, while the latter bets on three months later.

But in this round of the AI cycle, many things are starting to change: the primary market is becoming more and more like the secondary market, believing in certainty.

AI has taken most of the VC money. According to PitchBook, in 2025, the AI and large model sectors took 65.6% of the U.S. VC deal value. That is, without more than two-thirds of the increase in the number of companies, they took more than two-thirds of the money in the market.

Johnny, the initiator of Honghu Club whom I once interviewed, told me that he compared the differences between the past and now: "Before Alibaba went public in 2012 and 2013, fundraising in the primary market was not as smooth as people imagined. Everyone thought it was too expensive. When ByteDance reached a valuation of tens of billions of dollars, there wasn't such a rush as there is today. People thought they could see the ceiling." But today, the concentration of leading AI companies is very high, and everyone is flocking to them, even scrambling for shares. "This kind of concentration is something we haven't seen before."

AI really needs more money, otherwise it won't be able to make it to the IPO and will be out of the game.

They have high costs: training models, expanding business, and maintaining service capabilities all require computing power, and behind computing power are data centers, electricity, and infrastructure.

Once they fall behind in computing power and power generation facilities, it means a production bottleneck. Even if they have a leading model, it's difficult to turn the technological advantage into revenue growth.

In this context, obtaining more funds has almost become the common goal of leading AI companies. Those who fall behind will be out.

Why are VCs willing to invest in money - swallowing machines like Anthropic and OpenAI? This is the "certainty first" logic of PEs.

Since receiving strategic support from Microsoft, widely believed to be about $10 billion, in early 2023, OpenAI has quickly become one of the hottest assets in the global capital market.

In the past, the top primary market projects were only available to a few funds, and many individuals wanted to participate but it was not open to the public.

But in this round of financing in February this year, OpenAI raised $3 billion from individual investors. Individuals now have the opportunity to participate in primary market projects.

In the past two years, Anthropic has become the room that global capital most wants to squeeze into. Amazon has entered, Google has entered. Whether it's the primary or secondary market, money is pouring in.

Then Anthropic did something very interesting.

In mid - May, an official announcement stated that any stock transfer without the approval of the board of directors is invalid. SPVs won't work, and platforms like Forge and Hiive have no official authorization.

What does it mean?

When everyone wants to squeeze through the same door, the person inside the door can decide at any time who can come in and who can't. The rules are written by the winners.

So the question is: When capital no longer spreads its bets but focuses on a few winners, do these winners really have enough returns to meet everyone's expectations?

02 VCs Turn into CCs

This kind of concentration and consensus is unprecedented.

For at least the past year, the global market has stopped discussing who will be the next OpenAI and Anthropic. Instead, it's about "who can get the quotas of Anthropic and OpenAI" and "who can get on the shareholder list of the hottest projects."

When investors' discussions change from "Is it worth it?" to "Can I invest in it?", the process of value discovery has actually ended. For many late - entering institutions, what they face is no longer a startup project waiting to be verified, but a scarce asset widely recognized by the market.

When more and more funds flow to the same few companies, new trading demands also emerge. The market wants to get shares, old shareholders want to cash in their profits, and employees want to release their wealth in advance. After consensus is formed, liquidity begins to emerge naturally.

For many investors, what they are waiting for is no longer the IPO itself. Because the market has already completed the transactions before the IPO.

And an important prerequisite is that in both China and the United States, the IPO market has been sluggish in the past two years, and a large amount of money is stuck in the primary market.

In the past, the biggest feature of venture capital was the willingness to wait. After investing in a company, it would accompany the company's growth for five, seven, or even ten years and then exit through an IPO.

But today, more and more funds have to pay attention to exit efficiency.

For many investors, the biggest risk is not misjudging the company, but misjudging the exit time, Johnny said. "This in turn drives investors to prefer projects with a clearer listing path and more definite exit expectations."

People who really love primary market investment will still continue to layout and support the projects they are optimistic about. Johnny believes that the change does not occur at the investment concept level, but at the behavioral level.

In theory, investors still want to invest in good early - stage projects, but in reality, if you don't grab the opportunity, the valuation will double next month. How can you explain to your LPs that your judgment is correct?

Pursuing certainty is also a way to be responsible to LPs.

The DPI crisis is a dilemma faced by both China and the United States. Cambridge Associates, one of the world's largest institutional investment consulting companies, mentioned that in U.S. venture capital funds, the median DPI of most vintages (fund establishment years) since 2015 is below 1x.

Although there is no unified and public DPI database in China, the numbers that are circulating are more or less the same.

Based on the exit pressure mentioned above and the current DPI pressure, it's better to invest in some certain projects to better explain to LPs. (Who knows if it's the LPs who are pressuring the institutions to invest in that project? Who knows if the LPs are also desperate to get in?)

Venture Capital is no longer about taking risks, but about pursuing Certainty and Consensus. It's recommended that we rename it CC.

03 The Disappearance of Angels

The primary market investment style resembles the secondary market, and the secondary market investment is moving towards the primary market. Isn't this a two - way pursuit?

PitchBook data shows that in 2025, $194.7 billion in VC investments came from non - traditional investors, which is one of the highest levels in the past decade. These non - traditional investors include hedge funds, mutual funds, and crossover funds.

Typically, institutions like Tiger Global, Coatue, Altimeter, and D1, which used to mainly trade listed technology companies, have started to frequently appear in the primary market, participating in the financing of leading projects such as OpenAI, Anthropic, Scale AI, and Databricks.

Among them, Coatue Management is heavily invested in NVIDIA and Microsoft in the public market while also betting on OpenAI and Scale AI; Tiger Global Management, D1 Capital Partners and other typical technology hedge funds are also appearing more and more frequently on the cap tables of top - tier AI projects.

Moreover, although it's not something that only happened after AI, more and more institutions have a 50 - 50 relationship between their VC and secondary market businesses.

In fact, the performance of many VCs is now being propped up by the secondary market. Many VC investors can't get carry, and the most money they earn comes from stocks like NVIDIA and SK Hynix.

The logic of the primary market pursuing certainty and the secondary market chasing star projects may be rational, feasible, and rewarding for investors.

But what about for entrepreneurs? If your track is not labeled with AI and you don't tell a good story, you don't even have a chance to be noticed.

When we are discussing the integration of the primary and secondary markets today, looking back ten years from now, we may not be discussing a market change, but the end of an era.

What era? The end of the era when a founder, a garage, and an idea could leverage capital.

This article is from the WeChat official account "Non - standard Players UnDefined", written by "No Consensus". It is published by 36Kr with authorization.