According to PANews, last fall, OpenAI CEO Sam Altman was first dismissed by the board of directors, then dramatically won back his position, leading to a major reshuffle of the company's board members. On Tuesday, former OpenAI board member Helen Toner finally revealed part of the story. The AI researcher criticized Altman in an interview for his lack of transparency, which left the board in the dark about key decisions at OpenAI, leading to his dismissal.
Toner revealed that Altman repeatedly concealed information from the board and misrepresented the company's ongoing business, even lying to the board in some cases. For example, when OpenAI's flagship product ChatGPT made its debut in November 2022, the board members knew almost nothing about it. Toner found out about it through Twitter. Additionally, she pointed out that Altman claimed to be an independent board member of OpenAI with no financial interest in the company, but in reality, he had been hiding his ownership of the OpenAI startup fund.
What disturbed Toner more was that Altman provided inaccurate information to the board about the AI safety process and began lying in an attempt to drive her out of the board. This was the real reason for the collective concern of the board and the beginning of discussions about dismissing Altman. Toner emphasized that all four directors came to a conclusion that they could not trust anything Altman told them. This was completely unacceptable for a board that should be responsible for independently supervising the company. The purpose of the board is not to help the CEO raise more funds.
Previously, after the OpenAI team responsible for AI safety, Superalignment, was disbanded, OpenAI established a supervisory committee led by Sam Altman to assess the safety of its AI models. Reports suggest that the new safety committee marks OpenAI's transition from an undefined non-profit entity to a commercial entity.