Everyone keeps asking: “Are we in an AI bubble?” But as is often the case, I hear a different question, followed by recognition: “Wait—they picked up another round?”
Take Anthropic. In March it raised a $3.5 billion Series E at a valuation of $61.5 billion. Just six months later, in September, it entered into a $13 billion Series F round. New valuation: $183 billion.
OpenAI, the startup that ignited the AI boom with ChatGPT, remains the pacesetter, achieving an unprecedented $500 billion valuation in a tender offer last month. This is higher than the $300 billion valuation achieved during the March funding round and the $157 billion valuation achieved this year as a result of the October 2024 funding round.
In other words, in the 12 months between October 2024 and October 2025, OpenAI’s valuation increased by about $29 billion every month – about $1 billion per day.
It’s not just LLM giants. Further down the AI food chain (but still ranked high), recruiting startup Mercor raised its $100 million Series B in February at a $2 billion valuation — and then raised another $350 million by October as the company’s valuation reached $10 billion.
More than a dozen startups with rising valuations have raised two or more funding rounds this year, including Cursor, Reflection AI, OpenEvidence, Leela Sciences, Harmonic, Fall, Abridge, and Doppel. Some, such as Harvey and Databricks, are currently reported to be in their third round.
These valuation growth spurts, especially on the scale of billions and tens of billions of dollars, are extraordinary and raise many puzzling questions, starting with: Why is this happening? Is this phenomenon a reflection of the strength of these startups, or the unique business opportunities presented by the AI revolution, or both? And how healthy is this kind of thing – what risk are startups and the broader market taking by raising so much capital so quickly and increasing valuations so rapidly?
ghost of 2021
Some industry insiders have been heard explaining that there is more to the current phenomenon than just frothy market conditions. While the ZIRP, or zero interest rate policy era, which peaked in 2021, raised the stakes of startups in several back-to-back rounds (cybersecurity startup Viz was valued at $1.7 billion in a May 2021 round, and its valuation jumped to $6 billion when it raised $250 million in October), the underlying dynamics were completely different then (not least because ChatGPT had not yet launched).
Tom Bigala, founding partner of Bison Ventures, said he doesn’t think it’s anything like 2021, when “companies will raise a round … not because they’ve made any kind of real progress or achieved any technological or commercial milestone.” Bigala said investor enthusiasm was so high and capital flowed so freely that the perception of momentum was often enough to attract more than one round of capital in a year.
And for every successful Wiz, there were several startups in the ZIRP-era that raised two or more rounds within 12 months that have since struggled (such as grocery delivery app Joker, NFT marketplace OpenSea, and telehealth startup Cerebral).
Terence Rohan, managing director of Verna Fund, says today’s multi-round startups are demonstrating real business traction: “The revenue growth we are seeing in select companies is without precedent. In some cases, one could argue that we are dealing with a new phenotype of startup,” Rohan said via email.
Many of today’s high-flying AI startups are posting impressive numbers, even if we should be skeptical of the ARR at this point. You have young companies like vibe coding startup Lovable, which went from zero to $17 million in ARR in three months, and conversational AI startup Decagon, which hit “seven figures” in ARR in its first half. Cursor is probably the most famous of all: the developer-focused AI coding tool went from zero to $100 million in ARR in one year.
Aydin Senkut, founder and managing partner of Felicis Ventures, described the back-to-back funding as a sign of a high-velocity market where the cost of being wrong is higher than ever. “The rewards now go to those who identify and support these outliers early,” Senkut says, “because being in the wrong area or too late can not only reduce returns, but it can also nullify them.”
“The reward is great”
While widespread excitement around generative AI is fueling a series of funding rounds, startups pushing the boundaries in certain areas are among the biggest beneficiaries of the trend.
Cursor, the buzzy AI coding startup, ends 2024 with a healthy valuation of $2.6 billion. Its valuation increased to $10 billion in June 2025, when Cursor raised $900 million in funding. This month, Cursor announced it was now worth $29.3 billion after it raised an additional $2.3 billion in capital from investors including Accel, Thrive and Andreessen Horowitz.
Harvey, an AI startup aimed at the legal industry, raised a total of $600 million in two separate funding rounds within the first six months of 2025, taking its valuation to first $3 billion and then $5 billion. In October, multiple outlets, including Bloomberg and Forbes, reported that Harvey had raised another round of funding that gives the startup a valuation of $8 billion.
Each is representative of their respective fields: coding and legal AI are both booming right now. Legal AI company Norm AI raised $50 million from Blackstone in November — hot on the heels of a $48 million Series B raised in March. Similarly, in coding, Lovable raised its $15 million seed round in February, followed by a $200 million Series A by July at a valuation of $1.8 billion.
Healthcare and AI are also hot, with companies like OpenEvidence raising their July Series B to $210 million at a valuation of $2.5 billion, followed by another $200 million in October at a valuation of $6 billion. Abridge (last valued at $5.3 billion) and Hippocratic AI (last valued at $3.5 billion) also fall in this category.
Max Altman, co-founder and managing partner of Saga Ventures, says this trend isn’t just the result of enthusiastic startup investors throwing money around. For some startups, raising money quickly is becoming part of the strategic playbook – an effective means of fending off the competition.
“What these companies are doing is very cleverly eating the Earth for their competitors,” Altman said. Luck“The reward is so big now, so many people are chasing it, So, a really amazing strategy is to suck up all the capital, get the best funds to invest in your company so they don’t invest in your competitors, Stripe did that really early on, it was smart – you become this force of nature that is too big to fail,”
That said, this doesn’t mean that everyone attracting large-scale capital is waiting for a winner.
When the foundation is not set
If quickly picking up multiple rounds can be a strategic advantage, it can also become a dangerous liability. Or, as Jennifer Lee, general partner at Andreessen Horowitz, says, these back-to-back fundraisers can go right — and they can also go wrong.
“They are perfect when capital directly drives product market fit and execution,” Lee said via email. “For example, when the company uses new resources to expand infrastructure, improve models, or meet larger demand.”
So when do they go wrong?
“That’s when the focus becomes on raising money from the building before the foundation is ready,” Lee said.
Like a skyscraper built on unstable ground, startups that can’t support extremely high valuations risk a painful collapse. The valuations of some of the hyped AI startups could look unsustainable (perhaps even unsustainable) in the public markets, should the startup get that far. The resulting realignment manifests in a decline in the value of employees’ equity, creating talent retention and recruiting risks. Many of the biggest IPOs of 2025, such as Chime and Klarna, had decisive valuation cuts from their 2021 highs.
In private markets, rapid funding rounds mean that cap tables can become increasingly complex as founder stakes diminish. And then, perhaps, the biggest risk: That some of these highly funded startups end up with wild waste rates that they may not be able to claw back if times get tough and the capital dries up. This could lead to layoffs, or worse.
Ben Braverman, co-founder and managing partner of Altman’s Saga, said this is ultimately a story about both the concentration of capital in AI and how VCs evolve their strategies after 2021. Venture capital has always been about the Power Law – that the big winners keep winning big – but this has become especially true as VCs chase consensus favorites more than ever.
“The story so far in 2021 is a flight to quality on all sides of the market,” Braverman said. “It appears that VCs made the same decision in the last cycle: ‘We’re going to put most of our dollars into a few brand names that we really trust. And obviously, that will have its consequences.’
One of those consequences is that more capital than ever is flowing into a limited group of AI darlings. And while term sheets are being signed at a rapid pace today, even bullish investors acknowledge that, like any cycle, there will be winners and losers.
“In this type of environment, investors sometimes fall into a trap where they think every new AI model company is going to look like OpenAI or Anthropic,” said Bigala of Bison Ventures. Luck,
“They’re giving those businesses big valuations, and it’s an option value for companies that are going to be the next OpenAI or Anthropic,” Begala said. But, he says, “a lot of them won’t necessarily grow at those valuations…and you’ll certainly see some losses.”
<a href=