OpenAI, the San Francisco-based A.I. powerhouse now valued at $80 billion, operates by a unique structure where it is a nonprofit entity that runs a capped-profit subsidiary in which investors can buy equity. However, CEO Sam Altman may be looking to transition the organization into a fully for-profit one, The Information reported last month. The move would be unusual, however, as OpenAI has already simultaneously reaped the benefits of positive publicity from being a nonprofit while receiving significant investments that typically go into a for-profit company.
OpenAI was founded as a nonprofit research lab in 2015 by Altman, Elon Musk, and Ilya Sutskever, among others. Born out of concern that financial incentives could lead A.I. astray, OpenAI declared in a blog post published upon its founding, “Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return. Since our research is free from financial obligations, we can better focus on a positive human impact.”
OpenAI describes its existing structure as “a partnership between our original nonprofit and a new capped profit arm” on its website.
In 2019, OpenAI introduced a capped-profit arm. The company describes its structure as “a partnership between our original nonprofit and a new capped profit arm.” Finding that relying purely on donations made it difficult for the organization to stay competitive, this dual-model allowed OpenAI to raise money for its capital-intensive research while staying true to its nonprofit mission.
However, in the fine print, OpenAI reveals that the cap on returns for investors is an outstanding 100x. For context, the most prominent A.I. stock, Nvidia, has risen around 30 times in the last five years. OpenAI’s profit cap is so high that it might as well not exist.
At the center of the model transition is OpenAI’s board
OpenAI maintains that it is accountable to an independent nonprofit board, whose members own no equity in the company. However, observers began questioning who actually gets to call the shots at the company after its former board tried to fire Altman late last year. Microsoft (MSFT), the largest corporate investor behind OpenAI with a $13 billion stake, agreed to hire Altman within three days of his firing. Altman won his job back at OpenAI only days after, and surprisingly, Microsoft appeared to have encouraged it. This raises the question: in the fierce race for A.I. talent, why did Microsoft not try harder to retain Altman from re-joining its competitor, OpenAI?
“What we call OpenAI should be called Microsoft A.I. Microsoft controls OpenAI,” said NYU Professor Scott Galloway in an interview with Tech.Eu. (In March, Microsoft tapped Mustafa Suleyman, a co-founder of Google’s A.I. lab DeepMind, to lead a new unit called Microsoft A.I.) Microsoft holds a non-voting observer role on the board of OpenAI. On July 3, Apple, which in June announced a partnership with OpenAI, said its App Store chief Phil Schiller would receive a similar seat on the board.
It is unclear how OpenAI may transition to a for-profit model; it likely may involve doing away with its non-profit board that oversees the company. In a request for comment from Reuters, OpenAI said, “We remain focused on building A.I. that benefits everyone. The nonprofit is core to our mission and will continue to exist.”
OpenAI’s capped-profit model is rare, but its hybrid governance model has a long history of precedent. Food retailer Newman’s Own is a nonprofit that wholly owns for-profit distributor No Limit, which produces and sells all Newman’s Own products. In 2022, Patagonia’s founder donated 100 percent of the for-profit clothing brand’s voting shares to a nonprofit, making it another for-profit corporation owned by a nonprofit.