The disturbing chain of events at OpenAI during November suggest that those leading artificial intelligence (AI) cannot entirely be trusted with good behavior and strong governance. The new board of directors at OpenAI has a unique opportunity to demonstrate otherwise.

On November 17, 2023, the OpenAI board of directors (made up of only four members at the time) fired Sam Altman, co-founder and chief executive officer (CEO) of OpenAI. Soon after, OpenAI executives, employees and investors began pressuring the board of directors to reverse its decision. Less than two weeks later, but only after gaining concessions on an acceptable way forward, three of the four members resigned, and Mr. Altman was reappointed CEO. These stunning developments exposed many issues at OpenAI, including weak governance, conflicts of interest and lack of transparency. Things happen for a reason, and there is much to learn in this case.

What is OpenAI?

OpenAI describes itself as an artificial intelligence “research and deployment company.”¹ It was founded in 2015 as a nonprofit organization. Its stated mission is “to ensure that artificial general intelligence (AGI) – by which we mean highly autonomous systems that outperform humans at most economically valuable work – benefits all of humanity.” In 2019, OpenAI established a “capped for-profit” company called OpenAI Global. This company is controlled by OpenAI, which continues to be a nonprofit organization. However, OpenAI Global issues shares to investors, in order to fund the operations, research, development, and commercialization of artificial intelligence products developed by the OpenAI group. In return, investors benefit from the profitability of OpenAI Global products, based on a predetermined maximum profit, which is capped at 100 times the amount of investments.² Figure 1 illustrates the unusual and complex corporate and governance structure of OpenAI, as illustrated on its website. It is reported that Microsoft has a 49% ownership in OpenAI Global, while other investors and OpenAI employees own 49%, and the OpenAI nonprofit owns 2%.³ However, the effective control of OpenAI Global rests with the OpenAI nonprofit.

OpenAI Global is said to be “legally bound to pursue the nonprofit’s mission, and carry out that mission by engaging in research, development, commercialization and other core operations.” According to the OpenAI charter, the profit caps “limit the maximum financial returns to investors and employees to incentivize them to research, develop, and deploy AGI in a way that balances commerciality with safety and sustainability, rather than focusing on pure profit-maximization.”⁴ OpenAI is currently funded in large part by Microsoft, along with an array of venture capital investors. OpenAI employees also own shares of the for-profit entity.

Why was Sam Altman fired?

An official announcement on the OpenAI website reveals that the former board of directors no longer trusted Sam Altman. “Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities. The board no longer has confidence in his ability to continue leading OpenAI.”⁵ According to people familiar with the board’s thinking, “members had grown so untrusting of Altman that they felt it necessary to double check everything he told them.”⁶ The firing was apparently not the result of a single incident, but rather “a consistent, slow erosion of trust over time that made [board members] increasingly uneasy” with the leadership of Mr. Altman.⁷

In a memo to the OpenAI leadership, the board mentioned that Mr. Altman’s “behavior and lack of transparency had undermined the board’s ability to effectively supervise the company.” The board further added that “this decision is not about product safety or security, the pace of development or OpenAI’s finances. Bottom line, this was governance issue that lies at the heart of how this uniquely structured organization executes its responsibilities and advances its mission.”⁸ According to the Wall Street Journal, a person familiar with the board’s proceedings said “the group’s vote was rooted in worries that he [Mr. Altman] was trying to avoid any checks on his power at the company.”⁹ Another factor driving the board’s decision was “the board members’ lack of clarity around Altman’s pursuits outside of OpenAI.” There were “mounting concerns that OpenAI’s intellectual property and technology could be used in ways that made the board uncomfortable” people said.¹⁰

It was also reported that prior to his firing, Sam Altman clashed with Helen Toner (one of the board members who resigned). Mr. Altman apparently approached other members, trying to convince them to remove Ms. Toner. Some board members compared notes on their individual discussions with Mr. Altman. The members concluded that in one discussion with a board member, Mr. Altman left a misleading perception that another member agreed that Ms. Toner should leave. By this point, several members had concerns about Mr. Altman’s honesty.¹¹

Can Sam Altman be trusted?

Sadly, it is not the first time that Sam Altman has been dismissed for divided loyalties and lack of transparency. Four years ago, Paul Graham, the founder of Y Combinator (an American technology startup accelerator) fired Sam Altman over concerns that he was putting his personal interests ahead of Y Combinator interests. Over the years, it is said that Mr. Altman “developed a reputation for favoring personal priorities over official duties and for an absenteeism that rankled his peers and some of the start-ups he was supposed to nurture” said people on conditions of anonymity. “It was the school of loose management that is all about prioritizing what’s in it for me” said one observer.¹²

A person who has worked closely with Mr. Altman in the past “described a pattern of consistent and subtle manipulation, that sows division between individuals.”¹³ A former OpenAI employee said that Sam Altman lied to him on various occasions, and was “deceptive, manipulative, and worse to others.”¹⁴ The former employee said he was “disinclined to support Altman after working for him for two years.”¹⁵ When Mr. Altman was nice to people, it was mainly for personal reasons according to this former employee. In the months leading to his dismissal, Mr. Altman had been spending time exploring two business endeavors, and had spent weeks in the Middle East trying to raise money for one of them.¹⁶ These endeavors were apparently not part of his job responsibilities at OpenAI.

Why was Sam Altman reinstated?

Despite his apparent faults, Sam Altman has qualities that please many employees and investors, namely his “strategic skill sets, including his ability to be a matchmaker among powerful people.”¹⁷ People close to him describe his “uncanny knack for giving strategic advice, for negotiating business deals and for spotting undiscovered talent.”¹⁸ According to analysts, Mr. Altman’s “career arc speaks to the culture of Silicon Valley, where cults of personality and personal networks often take the place of stronger management guardrails.”¹⁹ At a recent OpenAI development day, “Mr. Altman presented as a millennial Steve Jobs, and announced plans for the company to become a dominant platform in generative AI.”²⁰  Analysts say that Mr. Altman “has used his shrewd maneuvering to stifle smaller open-source competitors, in this case to secure the future of his company employees.”²¹ “Sam lives on the edge of what other people will accept” mentioned a person who worked with him.²²

Employees and investors also have financial interests in keeping the OpenAI leadership stable to avoid delays in a planned sale of shares expected to earn almost $90 billion, representing a 300% increase of the OpenAI market value from earlier this year.²³ Many employees are expected to sell some or all of their shares at a significant profit. It is therefore not surprising that more than 95% of employees signed a letter threatening to leave if Sam Altman was not reinstated. Microsoft played a key role in support of Mr. Altman and his followers, by offering everyone a job and a similar working environment at Microsoft. According to analysts, “access to the [OpenAI] technology is important for Microsoft because it is underpinning many of its most important products and its future, as the company battles other tech giants in AI.”²⁴ The pressure by employees and investors forced the OpenAI board to resign or risk a potential collapse of the organization.

What is the way forward at OpenAI?

A collapse of OpenAI is not desirable, but neither is a lack of transparency or the pursuit of interests that are inconsistent with assigned mandates and job responsibilities. Senior executives must lead by example. The past and present behaviors of Sam Altman are troubling. A clear mandate letter, close supervision, and a trust but verify mantra are the way forward for the new OpenAI board when it comes to their relationship with Mr. Altman and the OpenAI executive team. It can be expected that Mr. Altman’s actions will be under close scrutiny. There should be zero tolerance for any kind of misconduct. Sam Altman may seem to come out on top, but he will face much greater oversight.²⁵

It is reassuring that the former board members did not give in to Mr. Altman’s wishes to become a member of the board. Separating the roles of chairperson and chief executive officer is a hallmark of good governance. In addition, the former board members appointed two new members with extensive experience, and a member of the original board kept his seat. The parties also “agreed to an independent investigation into Altman’s conduct and the board’s decision to fire him” people familiar with the matter said.²⁶ The new chairperson confirmed that this investigation will be led by external counsel.²⁷ The new board is also expected to appoint “as many as six additional members in addition to continuing its duties.”²⁸ Microsoft will have an “observer role” on the OpenAI board, namely a person without voting rights that will attend board meetings and report to Microsoft.

These are all positive developments. Governance will improve at OpenAI, provided that the new board members have the right character and motivations, and collectively have a relevant mix of knowledge and experience that complement one another, for the benefit of OpenAI and all of its stakeholders. Most importantly, board members must develop and maintain competencies and skills in matters of governance. The National Association of Corporate Directors (NACD) mentions that “effective onboarding of new directors is critical to success.” It provides its 24,000 members with “timely resources to help them navigate the challenges of the ever-changing business environment.” The NACD publishes a Guide for New Directors, which facilitates onboarding.²⁹

The development and applications of AI involve many risks. Some of these risks are known, while others remain unknown. AI is a highly disruptive and rapidly evolving technology that presents great opportunities, but also considerable threats that must be carefully managed. The stakes are very high, at a time when governments are scrambling to develop regulations. It is absolutely essential for organizations leading and implementing AI to adopt strong ethical and governance standards, and to be transparent with their actions in order to maintain public trust. The new board of directors at OpenAI has a unique opportunity to rise up to this challenge.

__________________________

¹  OpenAI website: https://openai.com/about (Accessed December 12, 2023).
²  Wikipedia website: https://en.wikipedia.org/wiki/OpenAI#cite_note-36 (Accessed December 12, 2023).
³  Anna Tong, Krystal Hu, Jeffrey Dastin, “OpenAI’s ousted CEO Altman in talks to return day after firing” The Globe and Mail (Nov. 20, 2023).
⁴  OpenAI website: https://openai.com/our-structure (Accessed December 13, 2023).
⁵  OpenAI website: https://openai.com/blog/openai-announces-leadership-transition (Accessed December 7, 2023).
⁶  Keach Hagey, Deepa Seetharaman and Berber Jin, “Behind the Scenes of Sam Altman’s Showdown at OpenAI” The Wall Street Journal (November 22, 2023).
⁷  Ibid.
⁸  Ibid.
⁹  Elizabeth Dwoskin, Nitasha Tiku, “Altman’s polarizing past hints at OpenAI board’s reason for firing him” Washington Post (Nov. 22, 2023).
¹⁰  Deepa Seetharaman, Berber Jin and Keach Hagey, “OpenAI Investors Keep Pushing for Sam Altman’s Return” The Wall Street Journal (November 21, 2023).
¹¹  Meghan Bobrowsky and Deepa Seetharaman, “The OpenAI Board Member Who Clashed With Sam Altman Shares Her Side” The Wall Street Journal (December 7, 2023).
¹²  Elizabeth Dwoskin, Nitasha Tiku, “Altman’s polarizing past hints at OpenAI board’s reason for firing him” Washington Post (Nov. 22, 2023).
¹³  Ibid.
¹⁴  Ibid.
¹⁵  Ibid.
¹⁶  Ibid.
¹⁷  Ibid.
¹⁸  Ibid.
¹⁹  Ibid.
²⁰  Ibid.
²¹  Ibid.
²²  Ibid.
²³  Deepa Seetharaman, Berber Jin and Keach Hagey, “OpenAI Investors Keep Pushing for Sam Altman’s Return” The Wall Street Journal (November 21, 2023).
²⁴  Keach Hagey, Deepa Seetharaman and Tom Dotan, “OpenAI Talks Continue as Sam Altman, Company Push to Reunite” The Wall Street Journal (November 21, 2023).
²⁵  Aditya Soni, “Altman’s return to OpenAI may come with fewer checks on his authority” The Globe and Mail (Nov. 23, 2023)
²⁶  Berber Jin, Deepa Seetharaman, Tom Dotan, “OpenAI Got Its CEO Back. What Happens Next?” The Wall Street Journal (Nov. 23, 2023)
²⁷  Newley Purnell and Keach Hagey, “Larry Summers Is OpenAI’s Surprise Pick to Mend Fences” The Wall Street Journal (November 22, 2023); and Deepa Seetharaman, “OpenAI Says Sam Altman to Return as CEO” The Wall Street Journal (November 22, 2023).
²⁸  Ibid.
²⁹  National Association of Corporate Directors (NACD), Onboarding for Success: Maximize the Impact of New Directors (Correspondance email issued by the NACD, December 13, 2023).