OpenAI will give its board veto power over any decisions made by CEO Sam Altman after he returns to the helm, Bloomberg and Business Insider reported.

Sam Altman, CEO of OpenAIPhoto: Steve Jennings/Getty images/Profimedia

The information was released Monday in OpenAI’s new guide to the security of artificial intelligence systems it is working on, a document titled “System Preparedness,” which describes “processes for monitoring, assessing, predicting, and protecting against catastrophic risks created by increasingly powerful models.” “.

The guidelines published by OpenAI come amid concerns that the Silicon Valley startup has taken a major technological leap in AI in a project the company has dubbed Q*.

Some tech sites have speculated that Altman’s firing by OpenAI’s former board of directors may have been related to the advance due to disagreements between him and other board members over the safe development of some of these powerful AI systems.

OpenAI is trying to mitigate concerns about the AI ​​systems it develops

But a week after Altman’s return as head of OpenAI, Microsoft chief Brad Smith dismissed the speculation, saying such advanced AI systems would “take years, if not decades” to develop.

“But I still think it’s time to focus on security,” he added. Smith made these comments because Microsoft is the largest sponsor of OpenAI.

As for OpenAI’s new security guidelines, they also call for the creation of an “Advisory Group” made up of representatives from multiple departments within the company to review reports on “emerging risks” and inform executive management and the board of directors.

But the document states that the authority to make final decisions will rest with the board of directors, not with Altman’s executive management.

“While management makes decisions, the Board of Directors reserves the right to overturn decisions,” the document states.

Sam Altman stated that he does not want absolute power in the company

Last month, OpenAI reshuffled its board of directors after the previous board fired Altman as CEO.

In June, Altman spoke about OpenAI’s leadership structure and his authority at the company, saying in an interview with Bloomberg that:

“You shouldn’t trust anyone here. I have no super voting shares. I don’t want them. The council can fire me. I think it’s important.”

Despite his comments since then, he said in a new interview this month that his ouster as CEO of OpenAI was one of the most painful moments of his life.