How did the highest managers and board members react?
There was plenty of leadership turmoil at OpenAI. The disagreements escalated in November 2023 when the corporate's board briefly fired its CEO Sam Altman. He got his job back in lower than every week, after which three board members resigned. The outgoing directors were proponents of constructing stronger guardrails and promoting regulations to guard humanity from potential harm from AI.
Since then, over a dozen senior employees have resigned, including several other co-founders and executives accountable for overseeing OpenAI's security policies and practices. At least two of them have joined Anthropic, a competitor founded by a former OpenAI executive accountable for AI security. Some of the departing executives say Altman pushed the corporate to bring products to market early.
Security “has taken a back seat to shiny products,” said Jan Leike, former head of OpenAI’s security team who left in May 2024.
Open AI CEO Sam Altman, center, speaks at an event in September 2024.
Bryan R. Smith/Pool Photo via AP
Why should the structure of OpenAI change?
OpenAI's wealthy investors cannot own shares within the organization or receive a seat on the board under the present nonprofit governance structure. This is because OpenAI is registered as a non-profit organization whose purpose is to learn society and never private interests. To date, all investment rounds, including a reported total of $13 billion from Microsoft, have been conducted through a for-profit subsidiary of the nonprofit.
The current structure allows OpenAI to simply accept money from private investors in exchange for a future portion of its profits. However, these investors don’t receive a voting seat on the board and their profits are “capped”. According to previously published information, OpenAI's original investors cannot earn greater than 100 times the cash they provided. The goal of this hybrid governance model is to balance profits with OpenAI's security-focused mission.
Converting to a for-profit company would allow its investors to accumulate ownership shares in OpenAI and not face a cap on their potential profits. Later, OpenAI could also go public and lift capital on the stock exchange.
According to a Bloomberg article that cited unnamed sources, Altman is reportedly in search of to personally acquire a 7 percent equity stake in OpenAI.
According to BoardSource, an association of nonprofit board members and executives, this arrangement will not be permitted for nonprofit executives. Instead, the association explains, nonprofits must “reinvest surpluses back into the organization and its tax-free purpose.”
What form of company could OpenAI change into?
The Washington Post and other media outlets, citing unnamed sources, have reported that OpenAI could change into a “non-profit corporation” – an organization aimed toward benefiting society and making a profit.
Examples of firms with this status, generally known as B Corps., include outdoor clothing and equipment company Patagonia and eyewear manufacturer Warby Parker.
According to the B Lab, a network that sets standards and offers certifications for B Corps, it's more typical for a for-profit company – not a nonprofit – to change into a charity. This is unusual for a non-profit organization, as non-profit governance already requires that these groups profit society.
The boards of directors of firms with this legal form are free to take note of the interests of society, the environment and other people who are usually not their shareholders, but this will not be required. The board should still decide to make profits its top priority and sacrifice its performance status to please its investors. Online craft marketplace Etsy did this in 2017, two years after becoming a publicly traded company.
In my view, any try to convert a nonprofit organization right into a nonprofit corporation is a transparent departure from specializing in the nonprofit's mission. And there's a risk that converting to a nonprofit would just be a ploy to disguise a shift toward specializing in revenue growth and investor profits.
Many legal scholars and other experts imagine that OpenAI won’t completely abandon its hybrid ownership model because of legal restrictions on the transfer of nonprofit assets into private hands.
But I believe OpenAI has a possible workaround: It could attempt to dilute the nonprofit's control by making it a minority shareholder in a brand new for-profit structure. This would effectively eliminate the nonprofit board's power to carry the corporation accountable. Such a move may lead to an investigation by the state attorney general's office and possibly the Internal Revenue Service.
What could occur if OpenAI becomes a for-profit company?
There is quite a bit at stake for society.
The potential harm that AI could cause is far-reaching, and a few of it’s already evident, akin to misleading political campaigns and bias in healthcare.
If OpenAI, an industry leader, focuses more on making profits than ensuring the security of AI, I imagine these dangers could change into even worse. Geoffrey Hinton, who won the Nobel Prize in Physics in 2024 for his research on artificial intelligence, warned that AI could worsen inequality by replacing “many household tasks.” He believes there may be a 50% likelihood “that we are going to must face the issue of AI attempting to take over humanity.”
And even when OpenAI were to retain board members for whom security is a top priority, the one common denominator among the many members of the brand new company board could be their commitment to protecting the interests of the corporate's shareholders, who would expect a profit. While such expectations are common on a for-profit board, they represent a conflict of interest on a nonprofit board, where the mission must come first and board members cannot derive financial profit from the organization's work.
The deal would undoubtedly please OpenAI's investors. But wouldn’t it be good for society? The purpose of nonprofit control of a for-profit subsidiary is to be sure that the profit doesn’t interfere with the nonprofit's mission. Without guardrails to be sure that the board is attempting to limit the harm to humanity brought on by AI, there could be little reason to stop the corporate from maximizing profits, whilst its chatbots and other AI products endanger society.
Regardless of what OpenAI does, most artificial intelligence firms are already for-profit firms. In my view, the one approach to address the potential harm is thru higher industry standards and regulations which might be starting to take shape.
The governor of California vetoed such a bill in September 2024 on the grounds that it will slow innovation – but I imagine a slowdown is precisely what it’s, given the hazards AI already poses to society is obligatory.
