OpenAI said Friday that as it transitions to a new for-profit structure in 2025, the company will create a nonprofit corporation to oversee commercial operations, lifting some of its nonprofit restrictions and allowing it to operate more like a high-growth startup .

“The hundreds of billions of dollars that major companies are now investing in AI development demonstrate what is truly needed for OpenAI to continue pursuing this mission,” OpenAI’s board wrote in the post. “We once again have to raise more capital than we imagined. Investors want to support us, but with this size of capital they need conventional equity and less structural standards.”

The pressure on OpenAI is tied to its $157 billion valuation, which the company has reached in the two years since the launch of its viral chatbot ChatGPT and the start of the generative artificial intelligence boom. OpenAI closed its latest $6.6 billion round in October and was preparing to compete aggressively with Elon Musk’s xAI MicrosoftGoogle, Amazon and Anthropic in a market expected to reach over $1 trillion in sales within a decade.

Developing the large language models that lie at the heart of ChatGPT and other generative AI products requires continued investment in powerful processors, most of which are provided by Nvidiaand cloud infrastructure, which OpenAI receives largely from top backer Microsoft.

OpenAI expects losses of about $5 billion this year on revenue of $3.7 billion, CNBC confirmed in September. These numbers are increasing rapidly.

By converting to a Delaware PBC “with common stock,” OpenAI says it will be able to pursue commercial operations while hiring staff for its nonprofit arm and allowing that wing to undertake nonprofit activities in health care, education and science.

The nonprofit will have a “significant interest” in the PBC “at a fair valuation determined by independent financial advisors,” OpenAI wrote.

OpenAI’s complicated structure as it exists today is the result of its founding as a non-profit organization in 2015. It was founded by CEO Sam Altman, Musk and others as a research laboratory focused on artificial general intelligence (AGI), which is a completely futuristic The concept was time.

In 2019, OpenAI wanted to move beyond its role as a pure research laboratory and hoped to function more like a startup. So it developed a so-called capped profit model, in which the nonprofit organization continues to control the entire company.

“Our current structure does not allow the board to directly consider the interests of those who would fund the mission and does not allow the nonprofit to easily do more than control the for-profit organization,” OpenAI wrote in the post Friday.

OpenAI added that the change would “allow us to raise the necessary capital on traditional terms like our competitors.”

OpenAI’s restructuring efforts face some major hurdles. The most significant is Musk, who is in the middle of a heated legal battle with Altman that could have significant implications for the company’s future.

In recent months, Musk has sued OpenAI, asking a court to block the company’s conversion from a nonprofit to a for-profit entity. In posts on , which should function as the company’s planned new structure.

In addition to the duel with Musk, OpenAI has faced a drain of senior talent, in part due to concerns that the company has focused on bringing commercial products to market at the expense of security.

At the end of September, Mira Murati, OpenAI’s chief technology officer, announced that she would be leaving the company after 6½ years. On the same day, research chief Bob McGrew and Barret Zoph, a research vice president, also announced their resignations. A month earlier, co-founder John Schulman said he was moving to rival startup Anthropic.

Altman said during a September interview at Italian Tech Week that the recent executive departures had nothing to do with the company’s potential restructuring: “We have been thinking about it – our board – independently for almost a year as we think about it, “What it takes to get to our next stage,” he said.

These weren’t the first departures of big names. In May, OpenAI co-founder Ilya Sutskever and former head of security Jan Leike announced their departures, with Leike also joining Anthropic.

Leike wrote in a social media post at the time that disagreements with leadership over the company’s priorities were the trigger for his decision.

“In recent years, safety culture and processes have taken a back seat to shiny products,” he wrote.

An employee who worked under Leike resigned shortly after him, writing on promises to do the right thing later.”