AI is coming and “zombie unicorns” are infesting Silicon Valley

AI is coming and “zombie unicorns” are infesting Silicon Valley.

At that time, these two companies were not yet listed, but they became unicorns with valuations of more than $1 billion.

However, when Musk's SpaceX set a record for the world's largest IPO in June this year, and members of the "trillion club" such as OpenAI and Anthropic were already waiting for the carnival outside the gates of the US stock market, many unicorns of that year were on the verge of death: Breslow canceled the four-day work week and unlimited paid leave for

employees, and the company laid off employees four times in four years; all the business models that Livney was distressed to establish were no longer applicable.

In the streets of Silicon Valley, these former darlings are dragging their "remnants" around - plummeting valuations, difficulty in financing, unbalanced growth, unable to IPO and unable to be acquired by giants. They have become "zombie unicorns" described by The Economist, wandering on Sand Hill Road.

Hundreds of billions of dollars could evaporate In 2021, Verbit, an AI-powered speech transcription and real-time subtitle platform, will be valued at US$2 billion. Today, Verbit is valued at only about US$300 million, and its revenue has plummeted from US$100 million in 2022 to about US$40 million today.

E-commerce payment company Bolt Financial raised US$355 million in Series E financing in 2022, with its valuation soaring to US$11 billion. Two years later, the secondary market repriced it: $300 million—a 97% reduction.

There is also the "Bored Ape" series of NFTs that have been snapped up by stars such as Justin Bieber - the parent company Yuga Labs won a US$450 million seed round in 2022 with a valuation of US$4 billion. Back then a monkey head could sell for millions of dollars.

Now, the NFT trend has receded, the price has plummeted by 90%, and even the monkeys can’t laugh anymore.

The database of Stanford University professor Ilya Strebulaev has verified the cruelty of the market: as of May 2026, 332 of the 1,900 unicorn companies have raised funds at a valuation no higher than their peak, and 212 of them have a valuation of less than US$1 billion.

As many as 383 companies have not disclosed any new financing plans in the past three years, and 41 of them have lost their "unicorn" status. The technology has not changed, but the trends have.

Sdebulaev’s data reveals a pattern: nearly half of the companies whose valuations have been cut in half completed financing in 2021, the year when the bubble was most inflated. Data provider Pitchbook expects investors will demand results starting late next year.

The combined net valuation reduction is expected to be between $500 billion and $1 trillion as companies valued below their historical peaks reprice, seek buyers or go bankrupt. The migration of “unicorns”: From e-commerce to software services and financial services to AI In just a few years, valuations have varied dramatically.

Why do these once glorious unicorns suddenly collapse? They rose in the wave of venture capital and enjoyed the dividends of business models such as DTC (direct-to-consumer), SaaS (software as a service), and industry traffic hype.

However, when they fell, they followed a similar internal path: frequent changes in executives, multiple property rights disputes, high labor costs, accusations of exaggerated product promotion and exaggerated technical capabilities... In short, it is difficult to live up to their reputation.

In April this year, during the latest round of layoffs at Bolt Financial, Breslow posted a message on internal Slack: "In the future, Bolt Financial will operate with a more streamlined organizational structure and will use AI as its core." This e-commerce payment company, once valued at tens of billions of dollars, now only has AI as a lifeline.

Livny still remembers that night: the day OpenAI announced the launch of ChatGPT, he stayed up all night. “This was the turning point in my life from being a genius who founded a unicorn company valued at US$2 billion to the time when the company was on the verge of bankruptcy and I was abandoned by the company,” he said.

The arguments of these founders all point to AI as the external cause of the collapse of the company's moat. Analysts believe that the rise and growth of “zombie unicorns” mostly occurred before the AI era, and now they are unsustainable due to their inability to completely transform and the gap in valuation.

The investment landscape of venture capital institutions and private credit has also shifted, becoming increasingly AI-oriented. "If you're not an AI-centric company, you need extremely strong performance numbers to raise money," said Mercury CEO Immad Akhund.

AI has reconstructed functions such as personalized recommendations and product traffic, making traditional e-commerce models like Bolt Financial that are purely platforms unsustainable; traditional artificial intelligence companies like Verbit, which essentially consume a lot of manpower, are facing the impact of generative AI; venture capital has also

abandoned the "dream logic" of temporary traffic blessings such as Yuga Labs, and instead invested in AI assets with core performance.

A reporter from the Shanghai Securities News combed through the Hurun Research Institute's global unicorn ranking list from 2019 to 2026, and found a clear migration route in terms of US market data: in the 2019 list, the number of unicorns in the e-commerce field ranked first; the number of SaaS and financial technology fields led the list from 2020 to 2025; in the 2026 list, the number of unicorns in the field of artificial intelligence jumped to first, reaching 132.

It is worth noting that among the 156 new U.S. unicorns on the 2026 list, more than 50 are from the field of artificial intelligence, accounting for the highest proportion. Many new SaaS companies are also essentially AI companies.

In terms of downgrade and exit (that is, the valuation falls below US$1 billion), starting from the 2023 list, the number of U.S. companies that have downgraded and exited has increased sharply, focusing on e-commerce, traditional technology, blockchain and other fields.

This phenomenon verifies the structural "K-shaped" differentiation of the capital market, that is, venture capital funds are highly concentrated in high-quality tracks such as embodied intelligence, computing power, and large models, while the financing space for traditional industries or edge startups that lack technology iteration capabilities has been greatly compressed and faces valuation reconstruction.

The former unicorns either transformed into real AI beasts, or slowly ossified in silence and became another "zombie" on the streets of Silicon Valley. Where do “zombie unicorns” go from here? The flow of capital creates the craze of the times.

Against the backdrop of heightened global uncertainty, capital is chasing more certain investment returns and growth expectations. The story of Amy Wu is a mirror.

The former head of FTX Ventures, who participated in the financing of Yuga Labs, asserted four years ago that "Yuga Labs will become the Disney of Web3 in the future." But with the collapse of SBF’s cryptocurrency empire, Yuga Labs’ dream of the metaverse has lost its way.

Nowadays, Amy Wu has moved to Menlo Ventures, and her checkable achievements have been concentrated in the fields of consumer, artificial intelligence and e-commerce technology, such as artificial intelligence research company Slingshot and AI-driven music generation platform Suno.

In August 2011, venture capitalist Marc Andreessen made a shocking statement that “software is eating the world.” BIP Capital stated that this view is still applicable today, but this time the subject "software" needs to be replaced with "artificial intelligence". How fast does it "swallow"?

Citi previously downgraded the ratings of six software companies in its research report, mainly due to structural challenges posed by the rapid release of AI models and the impact on traditional application software business models.

Ginger Chambless, head of research at JP Morgan Commercial Bank, said that the current valuation premium gap between AI and non-AI companies continues to widen, and VC funds are highly concentrated in the AI and machine learning industries with strong economic benefits.

According to PitchBook data, in the first quarter of 2026, the top five deals (involving OpenAI, Anthropic, xAI, Waymo and Databricks) accounted for nearly 75% of total venture capital investment. Thousands of companies are grabbing the remaining 25%.

Jeff Courey, a global industry analyst at Wellington Investment Management, believes that currently, investors are more inclined to look for companies that can achieve both growth and profit margin improvement, have relatively "AI safe" competitive barriers, and whose revenue growth can clearly benefit from AI innovation.

Judging from discussions on capital flows, the logic of venture capital for companies has shifted to focus on high-quality core assets and deterministic performance. Only those tracks that are in line with substantive industry trends such as AI can transcend historical cycles.

In the article "Zombie Unicorns Haunt Silicon Valley" published by The Economist, the author ironically stated at the end: Some venture capital institutions are hoping that if there are superstar IPOs this year, perhaps the market will be less harsh on those slightly bleak start-ups.

This dream is indeed very sweet, and now it is true that Musk has launched the largest IPO in history on the way to "chasing stars". But in the face of performance data, the market will not favor one over the other, and supervision will treat companies equally.

Therefore, it is time for those “zombie unicorns” who are still wandering the streets of Silicon Valley to ask themselves a question: If they are not at the forefront of the times, where the environment is unfavorable and people are at odds with each other—where is their own moat? (

#Stocks #Tesla #AI #Earnings #IPO