OpenAI, once a nonprofit founded in 2015 with the goal of developing artificial general intelligence (AGI) for the benefit of humanity, has completed its transformation into a for-profit entity. The company, known for its development of ChatGPT, has transitioned into a public benefit corporation—a form of for-profit organization committed to positive societal impact. This restructuring marks the end of a complex legal process and sets the stage for OpenAI to more effectively raise funds and capitalize on its AI technologies.
The change was facilitated by approval from Kathy Jennings, the Delaware Attorney General. Delaware is OpenAI’s place of incorporation, and Jennings’ approval marks a significant step forward, potentially ending over a year of negotiations, regulatory scrutiny, and public discussion regarding the organization’s governance and future. The restructuring also involves California, where OpenAI is headquartered, which, along with Delaware, had been investigating the proposed changes.
As part of the restructuring, OpenAI has reorganized its ownership, striking a new deal with Microsoft. Under this agreement, Microsoft now holds approximately a 27% stake in OpenAI’s new for-profit venture, valued around $100 billion, given the company’s estimated worth of $500 billion. This marks a significant development in the partnership between Microsoft and OpenAI, initially founded by high-profile tech figures including Elon Musk and Sam Altman.
Despite its move to a for-profit model, OpenAI remains under the technical control of its original nonprofit entity. OpenAI emphasizes that the nonprofit will maintain control over the newly formed for-profit branch. This structure aims to ensure that the nonprofit’s mission influences the broader corporate decisions of the for-profit OpenAI.
In terms of AGI, which OpenAI defines as highly autonomous systems capable of outperforming humans in most economically valuable tasks, the company has outlined new governance mechanisms. OpenAI’s board originally held the authority to declare the achievement of AGI, a point at which their partnership with Microsoft would be reevaluated. However, under the new structure, such a declaration will now need verification by an independent panel of experts. Microsoft retains rights to OpenAI’s confidential research methods until the AGI is verified or until 2030, whichever comes earlier, and retains some commercial rights to OpenAI products beyond the AGI stage.
Amid these structural changes, OpenAI has also introduced the OpenAI Foundation. This body, derived from its non-profit roots, is set to allocate $25 billion towards initiatives focusing on health, disease cure, and mitigating cybersecurity risks associated with AI. However, specifics regarding the timeline over which this funding will be disbursed were not disclosed.
Critics like Robert Weissman, co-president of Public Citizen, have expressed skepticism regarding the nonprofit’s control over the new for-profit entity. Weissman likened the scenario to that of a corporate foundation, questioning the extent to which the nonprofit can independently impose its values on the profit-driven objectives of the for-profit organization. He suggests that the control purported by the nonprofit over the for-profit entity might be more symbolic than practical, casting doubts on the effectiveness of the nonprofit in ensuring adherence to its original humanitarian mission.
This transformation of OpenAI from non-profit to for-profit raises pertinent questions about the future trajectory of AI development and its alignment with public good, especially at a time when AI technologies like ChatGPT are becoming increasingly influential in various sectors. The revised setup with Microsoft and the ongoing role of the newly formed OpenAI Foundation will be crucial in determining how OpenAI balances profit-oriented activities with its stated commitment to benefiting society at large.
Read the full post on theguardian.com
