AI Corporate Governance and Ben & Jerry's Risk (corpgov.law.harvard.edu)

A recent analysis titled "AI Corporate Governance and Ben & Jerry's Risk" reveals concerning governance conflicts in AI firms like OpenAI and Anthropic, particularly the risks associated with self-appointed "mission guardians." These guardians are tasked with prioritizing the societal benefits of AI over profits, but the paper argues that this setup can create unmanageable tensions. The analysis draws a parallel to the troubles faced by Ben & Jerry’s, where its mission guardians caused significant harm to investors and ultimately failed to uphold their social mission, culminating in a crisis that affected parent company Unilever’s market value. The paper highlights OpenAI’s tumultuous governance decisions in November 2023, when the board’s safety concerns led to the firing of CEO Sam Altman, resulting in a mass employee exodus threat and a temporary shutdown of investor interests. Although OpenAI has since restructured to mitigate these risks, it remains heavily controlled by a nonprofit, limiting accountability to investors. In contrast, Anthropic has incorporated a “kill switch” allowing stockholders to remove its mission-guarding directors, reducing the potential for misalignment between mission goals and investor interests. This analysis emphasizes that without proper governance mechanisms like those in Anthropic’s framework, future AI firms may replicate the failures showcased by OpenAI and Ben & Jerry’s.