Ahead of Race to IPO, OpenAI Discussed Spinning Out Robotics, Hardware Divisions
Company is considering an Alphabet-like structure for its portfolio of products, though no discussions are currently active
- Sam Altman proposed spinning out OpenAI’s robotics and consumer hardware divisions late last year.
- OpenAI is pivoting its business around a new “superapp” to win over coders and enterprise users.
- OpenAI started its consumer-hardware division last May by acquiring io for $6.5 billion in stock.
OpenAI Chief Executive Sam Altman discussed spinning out the company’s robotics and consumer-hardware divisions late last year, a move intended to give them more room to grow without weighing down the core business.
As part of the plan, the two companies would have been able to raise external funding and operate more independently. But it was rejected, in part because OpenAI concluded the new entities might have to remain consolidated on its balance sheet, according to people familiar with the matter.
The proposal offers a window into the difficult trade-offs OpenAI has to make as it races toward an IPO. For years, Altman greenlighted ambitious projects that went well beyond its popular chatbot, hoping to set a high bar for research and product development. But it is facing more pressure to clamp down on so-called “side quests” that aren’t helping it meaningfully boost revenue.
OpenAI is pivoting its business around building a new “superapp” to win over more coders and enterprise users after falling behind rival Anthropic. It recently missed some internal user and revenue targets, though the company has said its growth recently accelerated after the launch of a new AI model. The company also cut its video-generation tool Sora to help it free up more computing resources for its core products.
OpenAI could revive the idea of spinning out different business lines in the future, people familiar with the matter said. Under such a scenario, the company could create a holding company similar to Alphabet, which Google started in 2015 to separate its core search business from more speculative bets like Waymo, its self-driving car division, and Verily, its life-sciences division.
In its financial statements, Alphabet separates out revenue from cash cows—including Google’s search business, YouTube and its cloud division—from the operating losses of more speculative bets. The structure allows investors to weigh the performance of Google’s core business against the scope of the company’s longer-term investments separately.
Other tech giants have offered investors similar transparency without such structural changes. Meta Platforms separately has disclosed the tens of billions of dollars in losses from its metaverse division, while Microsoft breaks out the performance of businesses like its videogame unit and LinkedIn, all within their financial statements.
The robotics and consumer-hardware divisions operate separately from the rest of OpenAI, and report directly to Altman. Their work is carefully guarded, and some employees have described them as separate startups within the company.
OpenAI started its consumer-hardware division last May, when it paid $6.5 billion in stock to acquire io, an AI company led by former Apple designer Jony Ive, and hired its staff of roughly 55. Altman has teased the details of the new device under development, telling staff last year that it would be fully aware of a user’s surroundings, able to rest in one’s pocket, and be a third core device a person would put on a desk after a MacBook Pro and an iPhone.
OpenAI disclosed in a legal filing this year that the device won’t ship to customers before the end of February 2027.
The company has also been working on robotics for years, at one point training a humanlike robotic hand to solve the Rubik’s Cube. Last year, it announced a research partnership with the robot delivery service Coco Robotics, where Altman is also a personal investor.
“We’re trying to figure out how to be very successful at robotics,” Altman said on the Core Memory podcast last month. “If you could pick one thing to make the U.S. competitive at manufacturing and the world of atoms in general, you would say we need a lot of robots that can build a lot, lot more robots.”