In the latest development of the ongoing debate surrounding OpenAI’s future and the potential trajectory of artificial intelligence, a group of influential individuals is urging the Attorneys General of California and Delaware to prevent OpenAI from transforming its unique non-profit structure into a for-profit entity.
In a published on April 23, figures like “AI Godfather” Geoffrey Hinton, Harvard legal expert Lawrence Lessig, and former OpenAI researchers argue that this shift would fundamentally violate OpenAI’s original goals.
The letter states that the proposed changes would eliminate crucial safeguards, effectively giving control and profits from potentially groundbreaking technology to a for-profit entity obligated to prioritize shareholder returns. This comes as OpenAI faces pressure, as failing to restructure by year’s end could cost them $20 billion and hinder future fundraising.
OpenAI was established in 2015 as a non-profit, aiming to ensure that artificial general intelligence (AGI) benefits everyone, not just a select few. AGI, defined by OpenAI as systems surpassing human capabilities in economically valuable tasks, was recognized as potentially transformative but also risky, especially under for-profit control. In 2019, to attract investment for building AGI, OpenAI created a “capped-profit” subsidiary managed by the original non-profit, enabling them to raise over $60 billion and become a top startup. CEO Sam Altman testified to Congress in 2023 that this structure ensures focus on their long-term mission.
In December, OpenAI proposed dissolving this structure, turning its capped-profit arm into a public benefit corporation that would control operations. The original non-profit would relinquish direct control but become a well-funded foundation through equity in the new company, funding charitable work in science and education. OpenAI claims this allows them to “raise the necessary capital with conventional terms.” Investors in their recent $40 billion fundraising round can if the restructure doesn’t occur by year’s end.
An OpenAI spokesperson stated that the board is committed to strengthening the non-profit and ensuring the public benefits from AI. They added that the for-profit would be a public benefit corporation, like Anthropic and xAI, but with a non-profit backing, ensuring that the non-profit benefits as the for-profit grows.
Under the restructure, board members must consider OpenAI’s original mission, although it would be balanced against profits. UC Berkeley professor Russell, a signatory of the letter, explains that the non-profit has the power to shut down the company if it deviates from its mission, but this “off-switch” is being disabled.
Sunny Gandhi of Encode Justice argues that the for-profit status of OpenAI’s competitors is irrelevant. He likens it to a conservation non-profit becoming a logging company simply because other logging companies exist. He believes OpenAI should not prioritize competitiveness over its original mission, which their structure was meant to prevent.
The letter targets Attorneys General Rob Bonta of California and Kathy Jennings of Delaware strategically. Elon Musk’s attempt to block OpenAI’s conversion was denied due to his legal standing, but the judge acknowledged that the issue of breaching OpenAI’s charitable charter merited further consideration, expediting the trial. Unlike Musk, the Attorneys General have a clear legal interest.
California’s Attorney General Rob Bonta’s office is already investigating OpenAI’s plans, and Delaware Attorney General Kathy Jennings has previously signalled she to scrutinize any restructuring. Their actions may set a precedent for whether corporate governance structures can preserve a company’s ideals amidst the financial pressures of the AI boom.