OpenAI, the startup behind the viral chatbot ChatGPT that has helped create a fervor among investors for companies working on so-called generative A.I., has been in talks to raise additional capital from Microsoft, which had previously invested $1 billion in the startup in 2019 and could use the technology to supercharge a variety of its software products. It is unclear whether the $10 billion would be fully in cash or whether that figure includes the value of services Microsoft will provide as part of its partnership with OpenAI.

Microsoft’s potential investment is separate from efforts underway by a group of venture capitalists to scoop up shares from OpenAI employees and other insiders in a secondary market transaction that could value the company at around $29 billion. The Wall Street Journal first reported the valuation figure of the tender offer, and it was independently verified by Fortune. Terms are not finalized and could still change for both the Microsoft investment and the tender offer. 

Under the terms of the deal, Microsoft would be entitled to 75% of OpenAI’s profits until it earns back its initial investment, according to Semafor, which first reported the figure of the planned investment. After that threshold is reached, Microsoft would have a 49% stake in OpenAI with other investors taking another 49%, and OpenAI’s nonprofit parent getting 2%, the publication reported, without saying what stake Microsoft controls until that point.

Mira Murati, OpenAI’s chief technology officer, in an interview with Fortune, declined to comment on Microsoft’s possible investment into OpenAI or on other fundraising efforts. A Microsoft spokesperson also declined to comment.

OpenAI has emerged as the standard-bearer for generative A.I. technology, a powerful new class of tools that some have likened in significance to the advent of the web browser or the iPhone. Using large data models, generative A.I. systems can create highly polished original content—from lengthy written essays and movie scripts to digital art and music—based on just a few commands from a user. Even in a tightening market where investors are demanding profitability from high-valued startups, investors have been clamoring to invest in OpenAI, given the widespread potential use cases for the technology and room for upside. 

The San Francisco–based A.I. startup is projecting $1 billion in revenue by 2024, as was first reported by Reuters and independently verified by Fortune. OpenAI only generated approximately $35 million in revenue in 2022, sources say, and the company is heavily loss-making at the moment. One of the startup’s revenue streams is its integration with enterprises using its API. The company, which still considers itself primarily a research organization, has been focused on developing its technology and not on monetization. But some of the company’s A.I. models are being made available to business and government customers of Microsoft’s Azure cloud computing services. It is not clear how much of that money flows back to OpenAI.

The company has an unusual corporate structure. Initially founded as a nonprofit by a group of Silicon Valley movers and shakers who include the company’s current CEO, Sam Altman, as well as Elon Musk and fellow member of the PayPal mafia, Reid Hoffman, OpenAI has as its stated mission the creation of “artificial general intelligence” (or AGI). The company defines this as an artificial intelligence system that is able to perform most economically useful work as well as or better than a human can.

In 2019, the company converted from being a nonprofit into a “capped-profit corporation.” Investors will have the maximum returns they can realize capped at a set multiple of their initial investment on the premise that if the company succeeds in its mission of creating AGI, it will be the most valuable company on the planet by some order of magnitude. It has been widely reported that the first set of funders in OpenAI following this change in corporate structure have had their returns capped at about 100 times their investment.

Since Microsoft’s initial investment in 2019, OpenAI has released a string of products that have captured the public’s imagination. The DALL-E family of image generators, including DALL-E 2 released this past summer, can generate images, including photorealistic ones, from text prompts that a user writes. OpenAI’s GPT models can compose long passages of text—including computer code—that is often indistinguishable from what humans would write. 

ChatGPT, the chatbot interface that OpenAI debuted less than a month ago, can hold fluent and coherent dialogues on a vast range of subjects. It can answer questions on almost any topic. And a person can ask it to write almost anything—from business plans and movie scripts, to blog posts, poetry, and Linux code—and produce startling humanlike outputs. The only problem is that sometimes the answers ChatGPT provides are not factually accurate. 

According to The Information, Microsoft may use OpenAI’s technology to power its search engine, Bing, and challenge search giant rival Google. Microsoft currently uses some OpenAI technology in products like GitHub’s Copilot, a feature that automatically suggests the next line of code a programmer should write, and in Microsoft Power Apps, software that enables people to easily create mobile applications. Microsoft offers access to OpenAI’s GPT and DALL-E technology to customers of its Azure cloud computing platform as well. It’s believed that Microsoft is also looking to incorporate OpenAI’s language A.I. into Office applications such as Microsoft Word and PowerPoint.    

OpenAI is planning to use the capital from this fundraise for computing resources, research, and staff costs, sources tell Fortune. 

OpenAI’s other investors include Hoffman’s charitable foundation and Khosla Ventures. Last year, Sequoia Capital, Tiger Global Management, Bedrock Capital, and Andreessen Horowitz reportedly purchased shares from preexisting shareholders in a sale valuing the company at around $20 billion, according to The Information.

Meanwhile, control of the company and its intellectual property continues to reside with the board of the OpenAI nonprofit foundation. Its board members include Altman and Hoffman as well as company cofounders Greg Brockman, who is OpenAI’s chairman and president, and Ilya Sutskever, a legendary A.I. researcher who serves as the company’s chief scientist. Also on the board are Adam D’Angelo, an early Facebook employee who went on to found question-answering site Quora; tech entrepreneur Tasha McCauley; and Shivon Zilis, a former venture capitalist who is an executive at Musk’s Neuralink and whose infant twins were reportedly fathered by Musk. Musk himself, although he helped found OpenAI, is no longer involved in the company and does not serve on the nonprofit’s board.