OpenAI isn’t like most companies. The differences go deeper than its futuristic business, reflecting the organisation’s non-profit/capped-profit structure and altruistic mission, and limiting the extent to which any takeaways from its ongoing leadership crisis are applicable to other companies (including the publicly-traded issuers that Glass Lewis covers).
Yet in spite of those differences, the debacle surrounding CEO Sam Altman’s departure nonetheless shines a light on the value of oversight and governance — and how a board that lacks a broad range of relevant experience can materially change a company’s financial position, its relationship with employees, investors and other stakeholders, and potentially the course of its future.
Mission & Governance Structure
Rather than the typical three-pronged corporate “org chart” of a management team handling day-to-day operations, overseen by a board of directors, who in turn represent the interests of shareholders, OpenAI’s governance structure is more complicated. The board oversees a 501(c)(3) non-profit organization that owns an LLC that controls a holding company, co-owned with employees and others, which is itself the majority owner of the “capped-profit” OpenAI Global LLC.
Source: OpenAI
Nested holding companies aren’t unusual, but they typically serve to optimise a company’s tax situation. That in turn reflects the typical goal of a corporation: maximizing profits. By contrast, OpenAI’s website notes that the non-profit board’s “principal beneficiary is humanity, not OpenAI investors”. While the company does have investors, principally Microsoft, the board of the non-profit isn’t accountable to them. Their marginalized position in the org chart provides an indication of the unusual relationship at play — including a 100x cap on returns (maximum $100 for any $1 invested).
Investor Communications
Based on recent events, unlimited upside isn’t all that OpenAI’s investors gave up. Engagement between shareholders and companies is a critical facet of an effective governance structure, and keeping investors informed about strategic and leadership decisions is part of the board’s core functions. When a minority shareholder owns almost half of a business, they would typically expect to be involved in, or at least kept informed of, material developments. Yet Microsoft and other investors were reportedly caught unawares by the decision to fire Altman.
This indicates that OpenAI’s non-profit board didn’t think it was necessary to involve for-profit investors in its decision. Indeed, Microsoft CEO Satya Nadella has stated that the tech giant did not “have any relationship with the non-profit board which has the governance of [the for-profit entity],” implying that the board had overlooked Microsoft. Notably, when it came time to negotiate a potential return for Altman, those talks were reportedly led by Nadella, rather than a representative of OpenAI or its board – and ultimately resulted in Altman joining Microsoft, rather than rejoining OpenAI.
Board Skills & Experience
In recent years, the role of a public company board has evolved from primarily one of watchdog to strategic leadership. From investors’ perspective, having an appropriate mix of certain attributes— in particular skills, experience, diversity, and independence — is essential to ensure that the board as a whole can satisfactorily perform its oversight duty, have informed opinions on all topics relevant to the company and effectively advise management on important strategic decisions.
As a non-publicly traded, not-entirely-for-profit entity, OpenAI doesn’t fit that corporate model, so it’s perhaps unsurprising that its board doesn’t either. It is small (even smaller once Altman and former chair Brockman were recused from discussions), with a collective skillset and base of experience focused almost entirely on emerging technology risk. That narrow-but-deep composition of computer scientists, startup entrepreneurs and futurist philanthropists highly attuned to tech-related risk makes sense in the context of the company’s “research and deployment” mission and non-profit roots — but begins to look merely narrow, and potentially limited, in the context of the broader strategic considerations that come along with what the company has grown into since its ChatGPT tool was made available to the public, and the scope of what it could yet become.
Of course, it matters what OpenAI itself wants to be, and there’s a case that the board was fulfilling its duties by protecting the “safety first” mission. Yet even so, the board appears to have been oblivious to the motivations of several key groups of stakeholders – not just investors but also employees, many of whom who are reportedly ready to follow Altman out the door. One of the directors has already signalled his regret for the move.
Certainly, when your business is aimed at revolutionising the relationship between labor and capital by making most of society’s work redundant, you’re likely going to take a unique view of risk management. But an understanding of the existential societal risks that arise from moving forward too quickly with new tech doesn’t necessarily provide an understanding of the strategic and financial risks to your organisation arising from decisions that alienate employees and investors and bring the potential for material public reputational damage.