Meta asks California AG to block OpenAI’s transition from non-profit to for-profit

Hello General Bonta:

As a California company that develops Generative AI technology, Meta Platforms, Inc. (“Meta”) is very concerned about OpenAI’s attempt to end the non-profit status under which it was established in order to establish a for-profit business. We urge you to review this proposed transaction, including the nature and timing of any transfer of assets from OpenAI’s non-profit entity to other entities. Failure to hold OpenAI accountable for its choice to build as a non-profit organization may lead to the proliferation of similar charitable startups until they become profitable. Californians have a direct and urgent interest in stopping this behavior. All for-profit activities of OpenAI and its related organizations must be temporarily suspended to protect investors and consumers alike.

In 2015, OpenAI filed its first certificate of incorporation with the State of Delaware, which reads:

This Association shall be a non-profit organization organized exclusively for charitable and/or educational purposes within the meaning of section 501(c)(3) of the Internal Revenue Code of 1986, as amended, or a corresponding provision of any future Internal Revenue Service of the United States. the law. The specific purpose of this organization is to provide funds for research, development and dissemination of technologies related to artificial intelligence… This organization is not organized for the private benefit of any person… The property of this organization is irrevocably dedicated[se] and no part of the net income or property of this corporation shall ever inure to the benefit of any director, officer or member thereof or to the benefit of any private person.

OpenAI has reaffirmed this commitment on its website over the years:

Seeing that there is no clear path in the public sector, and given the success of other large projects in the private sector, [OpenAI] decided to pursue this project in independent ways bound by a strong commitment to philanthropy. [OpenAI] they initially believed that the 501(c)(3) would be the most efficient vehicle to facilitate the development of AGI that is safe and profitable in scope while remaining unencumbered by profit interests.

Using this nonprofit opportunity, OpenAI has raised billions of dollars from investors to further its mission. The company represents to the State of California and worldwide that it will be operated without the motivation of profit. Investors and the public rightly rely on that guarantee.

Now, OpenAI wants to change its status while keeping all the advantages that got it to the point it is today. That is wrong. OpenAI should not be allowed to break the law by taking and redistributing assets it has built as a charity and using them for potentially huge profits.

In addition, the proposed reform of OpenAI does not just represent the future, the potential abuse of the business form. And we would urge you to examine whether OpenAI’s past practices are consistent with its obligations as a non-profit organization – specifically whether it has improperly depleted the assets of the non-profit organization by distributing the assets to third-party companies.

OpenAI’s behavior could have seismic effects in Silicon Valley. If approved, OpenAI’s restructuring will represent a paradigm shift for technology startups; allowing this restructuring would only entice investors to set up organizations as non-profits, raise hundreds of millions of dollars in tax-free donations to support research and development, and then take on a for-profit status as their technology is commercially viable.

Indeed, if OpenAI’s new business model works, non-profit investors will receive the same benefits as those who invest in conventional for-profit companies while benefiting from tax breaks provided by the government and, ultimately, the public. That would distort the market by requiring any startup that wants to stay competitive to use the same playbook.

We understand that Elon Musk and Shivon Zilis are currently seeking to represent the public interest in the Musk v. Altman, No. 4:24-cv-04722-YGR (ND Cal.). While we can no longer urge your office to take direct action, we believe that Mr. Musk and Ms. Zilis are well qualified and well placed to represent the interests of the people of California on this issue. Their early, fundamental roles in the creation and operation of OpenAI and as former members of its Board put them in a position to understand better than anyone what OpenAI was intended to be and how its current behavior deviates from its charitable purpose.

Meta is committed to openness and transparency in the transformative field of AI. OpenAI’s philanthropic promise to develop safe and broadly beneficial AI without commercial pressure is important, and should be maintained. Given the high speed at which OpenAI is proceeding with its profitable transformation, this is a special case with an urgent need for action.

We appreciate your attention to our ideas and are happy to answer any questions you may have.

Sincerely,

Meta Platforms