With OpenAI seeking profits, activist seeks payback to the public
Jan 6, 2025

With OpenAI seeking profits, activist seeks payback to the public

HTML EMBED:
COPY
Robert Weissman, co-president of Public Citizen, argues that at least $30 billion of the AI developer's $150 billion valuation should be invested in safety, ethics and access. "I think they've abandoned" their original vision, he says.

A battle is brewing over the restructuring of OpenAI, the creator of pioneering artificial intelligence chatbot ChatGPT. It was founded as a nonprofit in 2015 with the goal of developing AI to benefit humanity, not investors. But advanced AI requires massive processing power, which gets expensive, feeding into the company’s decision to take on major investors. Recently, OpenAI unveiled a plan to transition into a for-profit public benefit corporation.

That plan has drawn objections from the likes of Elon Musk, Meta and Robert Weissman, co-president of consumer advocacy group Public Citizen, which urged California authorities to ensure that as OpenAI reorganizes, it will repay much of the benefits it received as a nonprofit.

The following is an edited transcript of Weissman’s conversation with Marketplace’s Meghan McCarty Carino.

Robert Weissman: Being a nonprofit enabled [OpenAI] to accept donations, and that was the model for the nonprofit, just like it’s the model for other nonprofits. They had donated cash and in-kind donations of computational power to do their work to develop this new technology. They didn’t have to pretend that they were going to be able to return investment to people. They were asking them to do it on a charitable basis, and they were able to raise a significant amount of money and resources through that approach.

Meghan McCarty Carino: So I understand in 2019, OpenAI kind of transitioned to a different structure so that it could take money from investors like Microsoft. But the business, a capped-profit business, has still been governed ostensibly by the nonprofit. Now, it is looking to transition into a for-profit public benefit corporation. What does the law say about what is required to make this kind of pivot?

Weissman: This is a very, very unusual thing, and exactly the pathway that OpenAI has taken, maybe no entity has ever done. But there is a history of nonprofits converting to for-profits, which is not exactly what OpenAI is doing, but it’s the core story of what OpenAI is doing. And to make that conversion, they’re going to have to receive effectively the OK of the attorneys general of Delaware, where the operation is incorporated, and California, where they’re registered and where they do business.

In the history of these kind of conversions is, if you’re going to take assets out of the nonprofit sector, you’ve got to pay back to the nonprofit sector the value of what you’re taking. If you run a charitable enterprise, you can’t suddenly just privatize it and make it your own. If you’re the executive director of a charity, you can’t just give it away to a corporation, or you can’t just turn your nonprofit into a for-profit corporation because you were able to succeed and evolve under the umbrella of nonprofit arrangements, tax-deductible support and so on. You’ve got to pay back to the charitable sector. The biggest precedent for these kind of conversions is the conversion of nonprofit Blue Cross health insurance companies into for-profit Blue Cross health care companies. That took place across the United States and in all the states where that happened, the for-profit entity was required to pay back into the nonprofit sector an amount equal in value to what they were privatizing. Typically, that money was then devoted to a health care charitable foundation. Many large health care charitable foundations still exist, including in California, from those kinds of conversions.

McCarty Carino: So what has OpenAI suggested about how it would make this transition, and what concerns you about their plan?

Weissman: Well, things have evolved quickly and very recently. They’ve now announced an intention to do this, even though it had been rumored for some time. What OpenAI says they’re going to do is spin off their for-profit affiliate. Right now, they have a nonprofit board that controls a for-profit affiliate, and the for-profit would pay to the nonprofit the value of what they’re taking, and that would be owned by the nonprofit in the form of stock in the new for-profit, independent OpenAI. So they’re proposing to make that payment to the nonprofit sector basically by paying themselves, which we think is not a good idea.

McCarty Carino: So you wrote a letter to California’s attorney general, Rob Bonta, in September making the case for OpenAI to pay at least $30 billion and share any artificial general intelligence tech basically with a charitable foundation, an independent charitable foundation. How did you come up with that?

Weissman: Well, we first started communicating with the attorneys general in California and also Delaware after the very strange and highly publicized board shakeup at OpenAI in November 2023, because what we saw happen there was the nonprofit board trying to exercise control over the for-profit affiliate and losing. The for-profit forces in and around OpenAI basically overwhelmed the nonprofit board, forced them out and replaced them with new people. At that point, it seemed to us, this entity, whatever it was, this combo entity, was no longer really working as a nonprofit. It effectively had been turned into a for-profit. And we started to say, well, if that’s the case, then they have a duty to pay to the nonprofit sector the value of what they’re taking out, just as was the case with these Blue Cross conversions. Now, a bit more than a year later, OpenAI is saying, yeah, that’s right, actually we don’t want to pretend to be a nonprofit anymore; we want to make that conversion. So in advance of this happening, we said, if there is going to be a conversion, if it’s actually going to formally do that, or if it’s going to be forced to do that — to recognize what’s effectively already happened — what is the value that they have to pay back?

Now they’ve evolved a very strange structure, not all of which is transparent. It’s not clear that OpenAI, the nonprofit, has many shares or has much share interest in OpenAI the for-profit. However, by the terms of the setting up of the whole operation, it does have control over OpenAI the for-profit. So we said, look, at least they’re owed the control premium, which on the low end is 20% of the value of an acquired company in most transactions that take place and in the stock market and regular acquisitions. Well, 20% of $150 billion, which is what OpenAI is now being valued at, gives you a low end of $30 billion. There’s a lot of reason to think the numbers should be higher, maybe much higher than that, but $30 billion, we think, is the baseline of what needs to be paid. And again, for us, it doesn’t work if OpenAI just pays itself, basically has the for-profit pay the nonprofit that is an affiliate and not truly independent. It has to go back to the independent charitable sector, which would likely mean one or more independent foundations that could actually advance the interests of developing AI in the public interest, advancing artificial intelligence safety and ethics concerns and figuring out how to provide greater access to people to the new technologies that are emerging.

McCarty Carino: So how do you see this potential pivot impacting OpenAI’s original vision to put humanity over profits in the quest for artificial intelligence?

Weissman: Well, I think they’ve abandoned that. I think they’ve abandoned that before this conversion, which is why, from our point of view, it seems like the genie is already out of the bottle. They say they’re converting, they want to convert to a public benefit corporation, which would have the ability to consider both profit and nonprofit interests. But in fact, what we’ve seen with OpenAI over the last year, really with the launch of the popular version of ChatGPT, is that it has been the least concerned about safety, the most aggressive in introducing new technologies of all the AI companies. So they introduced the technology that Google more or less had, but was scared to put on the market because of concerns about safety and perhaps because of liability, but was forced to quickly play catch-up after OpenAI moved forward. We’re seeing that again with OpenAI introducing technologies that have unbelievably high-quality artificial voice capacity, so really making it possible and likely for people to be fooled by human-sounding voices that will become very easily deployable over the internet. And that’s also a technology that Google had, analyzed, found to be too risky and wasn’t going to introduce on the market. But once one competitor does it, the others quickly follow behind. So from our point of view, OpenAI has left behind the idea of prioritizing safety and ethics and is really more interested in being the first mover. So the old Silicon Valley “Move fast and break things” motto appears to be what OpenAI has taken up, despite supposedly being a nonprofit and supposedly being an operation prioritizing the interests of humanity over any profit consideration.

McCarty Carino: What role could this independent charitable foundation play in that context?

Weissman: Well, depending on the size of the foundation, it could support all kinds of research and advocacy and education and access programs. It could support research and innovation designed to ensure safety. It could support smaller startups that were moving in a nonprofit capacity and committed to staying in a nonprofit capacity. It could pay more attention to safety on the one hand, and access to new technologies on the other. It could support efforts to make sure that lower-income people had access to new technologies as they become available. It could train new people in becoming programmers and developers. It could support advocacy to push back on the monopoly power, companies like OpenAI. There is an awful lot it could do, and ideally, actually would be more than one single foundation. At that scale of resources, I think it would be better spread among many. But no matter what, it could really become a powerful player to offset the unfortunate move we’re seeing to develop these really fascinating technologies way ahead of consideration for safety, ethics and access.

The future of this podcast starts with you.

Every day, the “Marketplace Tech” team demystifies the digital economy with stories that explore more than just Big Tech. We’re committed to covering topics that matter to you and the world around us, diving deep into how technology intersects with climate change, inequity, and disinformation.

As part of a nonprofit newsroom, we’re counting on listeners like you to keep this public service paywall-free and available to all.

Support “Marketplace Tech” in any amount today and become a partner in our mission.

The team

Daisy Palacios Senior Producer
Daniel Shin Producer
Jesús Alvarado Associate Producer
Rosie Hughes Assistant Producer