AI startups should brace for a wave of down rounds, investors say. Here’s how it might play out.

Buzzy AI startups are struggling to command the same valuations from a few years ago.

Down rounds and flat rounds for VC-backed startups have hit an all-time high this year since 2014.

And AI startups, which have been the darlings of the tech ecosystem since late 2022, aren’t exempt from the trend.

Case in point: buzzy startup Inflection AI raised $1.3 billion at a $4 billion valuation in 2023. When Microsoft paid $650 million to license Inflection AI’s software earlier this year — and simultaneously hired key figures from the startup — the price tag fell short of the unicorn valuation the AI upstart had initially commanded.

The Big Tech giant stopped short of officially acquiring Inflection AI, but that didn’t stop the UK’s competition regulator describing the acquihire as a “relevant merger situation.” The deal was seen by some investors as an omen for a wider phenomenon in the AI ecosystem — an impending wave of flat rounds and down rounds.

“In light of the current hype surrounding the sector, we’ve observed early signs that the AI market may be overheating, with the increase in flat rounds, down rounds, and acquihires being indicative of the challenges ahead,” Andreas Riegler, general partner at APEX Ventures, told B-17.

AI startups raked in nearly $50 billion globally from VCs in 2023, per Crunchbase. However, investor optimism appears to be fading in 2024. Startups are struggling to raise at the lofty valuations they did after the advent of ChatGPT-3, so many are looking at alternative funding arrangements.

As growth slows, a spate of AI juggernauts have recently cut their valuations or entered acquihire arrangements that significantly slash how much they are worth. Jasper AI slashed its internal valuation in 2023 as growth stymied, while the founding teams of Character.ai and Adept were absorbed by Google and Amazon, respectively.

“Investors went crazy for AI startups with zero revenue, at higher valuations, which will be an issue,” said Umesh Padval, managing director at Thomvest Ventures. “Acquihires like Inflection.ai, Adept, and Character.AI will accelerate in the next 12 months, and these startups will have to face the music,” he said.

“If they could be acquired for an equal amount of the last round, it’s great, but it’s not happening,” he added. “The fact that they’re selling means they’re not growing otherwise.”

Down rounds are on the rise

In 2022, 90.2% of AI startups in the US raised up rounds, while 6.5% raised down rounds, according to PitchBook data. In 2024, the number of startups raising up rounds dipped to 81.1% — and down rounds rose to 11.4%.

In 2021, the multiples in the market were at an all-time high. By 2023, startups that raised at inflated valuations could no longer justify the price tag — so had to resort to flat rounds, Padval said. Now, the AI ecosystem is in for a similar reckoning.

Many startups entering these arrangements are strapped for cash because they focus on building foundational models, a capital-intensive process.

“A lot of them hired really expensive engineers to try to solve this problem,” a London-based investor who was not authorized to speak on the record told B-17. “They also spend a lot of money trying to train these datasets, and what they realized is that now they’re running out cash without any sort of traction to prove — so they’ve burnt a lot of money.”

The key problem is that these AI startups operate like typical software companies, “with the caveat that the infrastructure serving the customer was more expensive than before, so margins were lower,” said Francesco Ricciuti, a deep tech VC at Runa Capital.

“Those that torch VC’s money to chase growth will crash mid-flight. They will die just like the quick commerce startups,” he added.

Secondaries are becoming an attractive option

Some VCs are reticent to back AI startups at the same lofty valuations they commanded in 2023 — so companies are thinking of inventive ways to raise capital.

“Startups aren’t going out publicly and raising a flat round — but many are opting for secondaries as a means to do this,” the London-based investor told B-17.

Secondary share sales have had a renewed importance in the last few years. Because of the scramble for AI talent, startups are starting to hire people with the promise of giving secondary sales to employees as a perk, Padval said.

“They’re doing this to retain them — it’s not a typical phenomenon, it’s an AI phenomenon,” he added.