What's At Stake

Amid mounting pressure from our coalition, OpenAI reversed course and preserved its nonprofit status, abandoning plans to fully convert to a for-profit. This was a win. But staying a nonprofit on paper doesn’t mean that OpenAI is following the rules in practice.

​

To maximize shareholder value, the Foundation risks violating nonprofit law. To prioritize public interest, it risks breaching its fiduciary duty to stakeholders. There’s also no confirmation that the Foundation received fair value for the charitable assets it transferred to the for-profit during the pre-IPO restructuring. 

​

AG Bonta approved OpenAI’s restructuring based on OpenAI’s promise that it would put proper conflict of interest guardrails in place. Evidence from the Musk trial confirms what we've long suspected. OpenAI didn't give regulators the full picture when it made its case for that structure. That leaves questions still unanswered, like:

​

• How will OpenAI avoid conflicts of interest and self-dealing among executives and board members who sit on both sides of the table?

• How can the OpenAI Foundation be expected to operate independently, and engage in programming that may undercut the for-profit’s commercial interests, when its board overlaps almost entirely with the for-profit’s?

• What happens when shareholder interests conflict with the public interest the Foundation is legally bound to serve?

​

The stakes are not just technical; they’re moral. AI could help us adapt to climate change, detect deadly disease, and give us more time back in our busy lives. But it's also exacerbating inequality, greenhouse gas emissions, surveillance, unemployment, and the erosion of democracy. As OpenAI barrels toward an IPO that will make course-correcting far more costly, the public should benefit, not lose out, from a multi-billion dollar investment in AI meant for the public good.