OpenAI's Shifting Governance and Political Engagement

Key Insight

Following an internal 'blip' and subsequent investigation, it became clear that OpenAI's original nonprofit structure was unsustainable, especially as the company spent billions more than it earned annually. Altman set about transforming OpenAI, aiming for a less 'weird' structure, starting by removing certain internal influences. The company recruited experienced board members, including former Sony Entertainment president Nicole Seligman and former Bill and Melinda Gates Foundation CEO Dr. Sue Desmond-Hellmann, and implemented a new conflicts of interest policy to prevent issues seen previously.

Altman actively engaged in US politics, expressing alarm about President Biden's ability to win re-election against Donald Trump. Through his political advisors, the Krisiloff brothers, Altman funded polling and focus group research indicating voter skepticism about Biden's age, which was shared with Democratic challenger Dean Phillips. Despite pledging money and support to Phillips and hosting him, Altman ultimately did not endorse any candidate, instead tweeting a general endorsement of a 'reasonable, centrist candidate' focusing on economy, safety, and generational change, much to the annoyance of Phillips' campaign manager who noted Altman's 'penchant for secrecy.'

In the realm of AI regulation, OpenAI heavily influenced President Biden's October 30, 2023 executive order, with Altman visiting the White House four times that year. However, OpenAI's initial silence and subsequent tepid approval warned against slowing innovation, contrasting with the order's deep influence from Effective Altruism-aligned groups like the RAND Corporation, which pushed for provisions requiring detailed model development information. OpenAI, through lobbying firms, also successfully opposed a California AI safety bill that would have held tech companies liable for AI harms, with key figures like Ron Conway and House Speaker Nancy Pelosi contributing to its veto.