OpenAI Seeks Investigation into AI Voting Measures

OpenAI, under the leadership of Sam Altman, has reached out to California’s primary political finance regulator, urging an investigation into individuals behind AI-related voting initiatives. The company expressed concerns about the motives driving these initiatives.

The California Fair Political Practices Commission (FPPC) complaint mentions Alexander Oldham, who has proposed two pending measures. If approved, these would empower state regulators to oversee major AI firms, especially focusing on public interest corporations—something OpenAI recently transformed into.

Earlier this month, it was reported that Oldham is related to Zoe Blumenfeld, a senior member at Anthropic, OpenAI’s main competitor. He also knows Guy Lavigne, a tech entrepreneur involved in a contentious legal dispute with OpenAI regarding the company’s original concept.

It’s important to note that there’s no evidence linking Lavigne to the ballot initiative, nor is he mentioned in OpenAI’s claims.

OpenAI’s lawyers indicated in the complaint, which was reviewed by the media, that Oldham’s actions seem aimed at placing excessive regulatory burdens on them. They also claimed that Oldham may have breached state lobbying regulations by not disclosing necessary information.

“Experts suggest that the language within the initiative specifically targets OpenAI’s unique structure as a public benefit corporation. This could allow regulators to focus on individual companies rather than establishing standards for the industry as a whole. Additionally, Oldham’s connections to individuals with whom OpenAI has had ongoing conflicts raise serious doubts about the true motivations behind this initiative,” the complaint elaborated.

Interestingly, Oldham had no prior involvement in AI policy discussions or political campaigning before this ballot suggestion.

OpenAI’s representation posits that Oldham might merely be a front for the actual supporters of the bill, pressing for an inquiry into any ties to the Coalition for AI Nonprofit Integrity (CANI).

CANI supports another ballot measure proposed by Poornima Rama Rao, a former OpenAI employee turned whistleblower, which seeks to reverse changes at OpenAI.

OpenAI contends that the three measures show notable similarities, hinting they were possibly crafted by the same individual.

The media has found no evidence connecting Oldham to CANI.

Previously, OpenAI accused CANI of concealing funding and violating state lobbying rules. They allege that CANI could be a front group for Elon Musk, who is embroiled in legal action against OpenAI over its departure from a nonprofit model.

Last fall, the FPPC dismissed an earlier complaint from OpenAI against CANI, citing insufficient evidence for finance violation claims.

Notably, OpenAI hasn’t criticized Anthropic’s involvement in this situation.

In the recent complaint, similarities were drawn between Oldham’s relatively low-profile background and that of Jeffrey Mark Gardner, a past chair of CANI. Gardner previously led a nonprofit in New York without clear ties to California or the AI sector, and has now resigned.

OpenAI’s law firm stated, “When major political actions are taken through obscure entities, it diminishes public trust and clouds informed decision-making. We are asking the FPPC for greater transparency so the public can fairly evaluate these initiatives.”

This month, the California Attorney General’s Office provided a title and summary for Oldham’s ballot proposal, allowing him to start gathering the necessary 500,000 signatures for the fall ballot.

The FPPC, CANI, and Oldham haven’t commented on OpenAI’s filing yet.

Oldham mentioned using an AI chatbot to craft the ballot measure, intending to spark necessary discussions around AI regulation. He dismissed any collaboration with others, including his legal team.

He also stated, “To clarify, neither Guy Lavigne nor Zoe Blumenfeld were part of this initiative. I haven’t spoken to Guy in nearly a decade and Zoe in over two years. This initiative was solely created and funded by me.”

Anthropic similarly stated they have no involvement or knowledge of Oldham’s proposals.

Lavigne denied any collusion with Oldham or prior knowledge of the voting measures, emphasizing their connection is minimal at best.

He also remarked on their financial constraints, stating, “We don’t have the resources to support our voting efforts.”