- A former OpenAI board member wrote in The Economist that governments need to regulate AI companies.
- Helen Toner and Tasha McCauley, the only women on the company’s board of directors, stepped down in November.
- The two men supported the firing of OpenAI CEO Sam Altman, who they say created a “toxic” work environment.
Two former OpenAI board members join artificial intelligence company They cannot be trusted to govern themselves and need third-party regulation to hold them accountable.
Helen Tonner and Tasha McCauley had served on OpenAI’s board of directors until they resigned in November in a chaotic attempt to oust OpenAI co-founder Sam Altman, who swiftly returned as CEO just days after his firing and returned to the board five months later.
In an op-ed in The Economist, the two former directors cited senior executives who said the co-founder had created a “toxic culture of lies” and engaged in “inappropriate behavior,” and said they stood by their decision to fire Altman. [that] This could be considered psychological abuse.”
Since Altman returned to the board in March, OpenAI has faced questions about its safety efforts and has been criticized for using an AI voice in Chat GPT-4o that sounds eerily similar to actress Scarlett Johansson.
With Altman back at the helm, Tonner and McCauley write that OpenAI cannot be trusted to hold itself accountable.
“We also feel that developments since his return to the company, including his return to the board and the departure of safety-conscious senior personnel, bode ill for OpenAI’s experiment in self-governance,” they wrote.
For OpenAI to achieve its stated mission of benefiting “all of humanity,” Tonner and McCauley argued, governments need to step in and establish “an effective regulatory framework, now.”
The former directors wrote that while they once believed OpenAI could govern itself, “experience leads us to believe that self-governance cannot reliably withstand profit-driven pressures.”
OpenAI, Toner and McCauley did not immediately respond to BI’s requests for comment.
Policymakers must ‘act independently’ from AI companies
Toner and McCauley qualified their arguments for government regulation by acknowledging that poorly written laws can burden small businesses and stifle “competition and innovation.”
“It is crucial that policymakers act independently of large AI companies when crafting new rules,” the researchers wrote. “They need to be vigilant against loopholes, regulatory ‘moats’ that protect first-mover companies from competition, and the possibility of regulatory capture.”
In April, the Department of Homeland Security announced the creation of an Artificial Intelligence Safety and Security Commission to make recommendations for the “safe and secure development and deployment of AI” across America’s critical infrastructure.
The board’s 22 members include Altman and CEOs of major technology companies, including Nvidia CEO Jensen Huang and Alphabet CEO Sundar Pichai.
While the safety committee includes representatives from tech nonprofits, it is overrepresented by leaders of for-profit companies.
AI ethicists who spoke to Ars Technica expressed concern that undue influence from for-profit companies could result in policies that prioritize industry over human safety.
“If we can all agree that we’re interested in keeping people ‘safe’ when it comes to how AI is used, then I think we can also agree that it’s important to have people at the table who are dedicated to putting people over technology,” Margaret Mitchell, an AI ethics expert at Hugging Face, told Ars Technica.
A DHS spokesman did not respond to a request for comment.
Business Insider’s parent company, Axel Springer, has a global deal to allow OpenAI to train models on its media brands’ reporting.
