AI Ethics
21h ago
Harvard Law Paper Analyzes Governance Risks in AI Companies Ahead of Anthropic IPO
Jun 1, 2026
AI Summary
A new paper from Harvard Law examines the governance structures of AI companies, particularly Anthropic and OpenAI, highlighting the risks associated with investor-overriding mission guardians. The authors argue that such structures can harm investors and undermine the intended social missions, using the Ben & Jerry's case as a cautionary example.
- The Harvard Law paper titled 'AI Corporate Governance and Ben & Jerry’s Risk' discusses governance structures in AI companies, focusing on Anthropic and OpenAI as they approach public markets.
- The paper introduces the concept of 'Ben & Jerry’s risk,' which refers to the potential for mission guardians to negatively impact investors and fail to achieve their intended goals.
- The authors cite the example of Ben & Jerry's, which faced significant market value loss and governance issues after being acquired by Unilever, illustrating the pitfalls of such governance arrangements.
- OpenAI's governance structure, which included nonprofit directors overseeing a for-profit subsidiary, faced challenges, including a controversial board decision that led to CEO Sam Altman's temporary firing.
- OpenAI's conversion to a public benefit corporation in 2025 retained some mission governance features, but concerns remain about its long-term effectiveness and the potential for future modifications.
- Anthropic has filed for an IPO and employs a governance model that includes a 'kill switch' allowing investors to remove mission guardians, which the authors argue may provide a more workable solution.
- The paper concludes that of the three companies with investor-overriding mission guardians, two have faced significant failures, raising questions about the viability of such governance structures in the long term.
anthropicinvestor influencemission guardianscorporate ethicssafety mission