AI startups raised billions of dollars last year. Now many are struggling to survive—and asking Silicon Valley’s biggest companies to bail them out. Hiring a startup’s key employees and getting its technology in exchange for a licensing fee gives buyers most of what they would want from a purchase without having to get an OK from the government . More exits are coming, as a bubble built by the excitement around gen AI is peaking. Creating the models that power gen AI requires hundreds of millions of dollars in investment before returning a cent of revenue. Many startups are discovering they don’t have the resources to get there.“These companies know being able to buy hundreds of smaller firms without attracting challenges are over,”
Artificial intelligence startups raised billions of dollars last year, aiming to become winners in the latest tech-driven boom. Now many are struggling to survive—and asking Silicon Valley’s biggest companies to bail them out.
At least three once-hot AI startups have been rescued via a new type of deal that many in the tech industry say are acquisitions in everything but name. These deals have the advantage of skirting the typical regulatory process at a time when big tech’s growing control over generative AI is being scrutinized by governments.
On Friday, Character.AI announced a deal for Google to use its technology and hire many of its researchers and executives, including its co-founders Noam Shazeer and Daniel De Freitas. Google negotiated a licensing fee worth $2 billion for the startup’s technology to help buy out early investors, people familiar with the matter said.
The two companies considered an outright acquisition, but concluded that was unlikely to get past regulators, according to a person familiar with the matter.
In June, Adept AI struck a deal in which Amazon agreed to hire most of the startup’s employees and paying about $330 million to license its technology, according to people with knowledge of the arrangement. That was enough, along with Adept’s remaining cash, to pay back investors, but a disappointing turn for a company that just last year was valued at $1 billion.
Microsoft cast the mold for this deal type in March when it hired nearly all the employees from AI developer Inflection to start a new consumer AI division and paid around $650 million to license its technology.
More exits—either pseudo-acquisitions or real ones—are coming, investors say, as a bubble built by the excitement around generative AI is showing signs of peaking. Creating the large-language models that power generative AI often requires hundreds of millions of dollars in upfront investment before returning a cent of revenue. Many startups are discovering they don’t have the resources and runway to get there.
“There were a lot of companies that raised on a big vision, but not tangible examples and actual detail,” said Shaun Johnson, a founding partner at the AI-focused venture firm AIX Ventures.
A bailout by a tech giant is better than a flameout, but far from the disruption of the tech industry’s status quo investors were betting on. Instead, the deals have further strengthened the tech giants, which can foot enormous AI bills with their cloud computing infrastructure and existing revenue machines.
Growing skepticism for AI
Anxiety about the outlook for the AI boom has contributed to a tech swoon in the public market recently that has knocked about 13% off the tech-heavy Nasdaq Composite over the past month.
The growing industry skepticism is a marked change from last year, when money poured into AI startups following the blockbuster debut of OpenAI’s ChatGPT in November 2022.
That same year, chief executive David Luan launched Adept to create AI agents that could handle office tasks like paying invoices or processing insurance claims.
Adept’s agents ran on specially built large-language models—an undertaking that can cost upward of $100 million.
Adept raised a little more than $400 million, but the costs of building its technology exceeded what its founders had anticipated, according to a person familiar with the matter.
This spring, Luan and his team reached out to companies including Microsoft and Salesforce, asking if they would buy it, according to people familiar with the matter.
Adept ultimately struck a deal with Amazon, where Luan and some 50 other Adept employees have taken leadership positions on a recently formed team tasked with helping the e-commerce giant catch up with its peers in the AI arms race.
About 25 Adept staffers are staying at a scaled-down version of the startup that won’t create new AI models, but will use and sell its existing technology.
New role for Character
Character followed a similar trajectory. It created AI-powered chatbots that can simulate anyone from a psychologist to an anime character to Elon Musk. Many people used the service for romantic role-play, which the company discouraged.
Last year the company raised $150 million in a deal that valued it at $1 billion. But it struggled to generate enough revenue from paid users and by this summer was talking to Facebook-parent Meta about a bailout, according to people familiar with the matter.
Ultimately, Character homed in on Google, where Shazeer used to work, and where he co-wrote a seminal 2017 paper that laid the groundwork for much of the generative AI technology in use today. Shazeer, Character’s chief executive, is still highly regarded at Google and is close with co-founder Sergey Brin, who helped seal the deal.
Google’s licensing fee comes on top of an earlier $500 million investment the tech giant made in Character. The money from Google will be used to buy out early shareholders at a valuation of $2.5 billion—or more than twice the level last year—and to fund the startup’s growth.
Shazeer and about a quarter of Character’s employees are joining Google’s AI research division to work on its flagship Gemini systems. The rest will remain at Character to continue working on chatbots and will use more open-source AI models—a cheaper approach than relying solely on their own.
The Information previously reported Character’s new valuation in the Google deal.
Regulator Scrutiny
The Biden administration’s increased actions to block technology mergers and acquisitions are one reason for the unusual structure of the Character, Adept and Inflection deals, according to people in the industry.
“These companies know the days of them being able to buy hundreds upon hundreds of smaller firms without attracting challenges are over,” John Newman, a law professor at the University Miami who focuses on antitrust and competition, said of tech giants like Amazon and Google.
So-called acqui-hires, in which a large company buys a startup mostly to hire its employees, have long been common in Silicon Valley. But hiring a startup’s key employees and getting its technology in exchange for a licensing fee is unusual. The structure gives buyers most of what they would want from a purchase without having to get an OK from the government.
The Federal Trade Commission is probing both Amazon’s deal with Adept and Microsoft’s with Inflection to see whether either buyer structured the arrangement to avoid government approval, people familiar with the matter said.
0 comments:
Post a Comment