Sam Altman Tells OpenAI Staff Investigation Will ‘Soon’ Close, Employees Await Surprises
Inside OpenAI, Sam Altman is still approachable and friendly to staff, even though many of them are realizing how little they know about how the company operates.
In a recent all-hands meeting, which the CEO hosts regularly allowing employees to ask questions in real time, a staffer asked about the internal investigation into Altman that was launched after his dramatic ouster and return late last year. Altman responded simply that it would be “over soon,” supplying staff with no other detail, according to two people who witnessed the exchange.
The expectation internally is that the investigation will be completed before the first quarter is up, one of the people added. The internal investigation nearing its end was also reported by the Wall Street Journal and The New York Times. Altman addressing it with employees was not previously reported.
Such a lack of candor with employees is a more recent development for Altman, one of the people familiar noted. His public profile has risen since being pushed out last year of the company he co-founded by his own board of directors, a period now referred to by many OpenAI employees as “the coup,” and increased scrutiny has come along with it. While Altman is still willing to chat in the hallways and will respond to any employee on Slack, there are many core elements of the operation that remain a mystery to employees.
A recent revelation to the SEC that Altman is the sole owner of the OpenAI Startup Fund came as a surprise to employees because it was never disclosed internally. “No one knew that,” one of the people familiar said. An SEC investigation into Altman’s communications, as reported in the Wall Street Journal, was another recent surprise. Some employees are concerned with what the result of such a probe might be. There is also concern over the FTC’s investigation into OpenAI and its ties with Microsoft.
While most employees prefer to keep their heads down and work, many of whom are actively working on GPT-5, the next version of OpenAI’s generative AI system set to be released in the coming months, there is “a lot of pressure” that’s come with increased public awareness of Altman, one of the people familiar said.
Many employees are also confused, as are many observers, about the actual operational structure of OpenAI. It has a nonprofit arm in control of a for-profit arm, which itself has control of a holding company, which controls another for-profit entity, and that entity is what Microsoft is technically invested in.
“It doesn’t make a lot of sense to anyone,” one of the people familiar said of the company’s structure. The person wondered if the various investigations would force some kind of change in the next year or so, perhaps simplifying the operational structure.
“I’m expecting some more drama, from somewhere,” the person added.
OpenAI did not respond to a request for comment.
Another opaque matter is the form of equity compensation employees receive in the form of profit participation units, or PPUs, which make up a large portion of their expected pay. Employees are still not told how many PPUs are included in their grants, only the assigned dollar value. Although OpenAI employees are paid well, many workers are increasingly concerned about how much risk they’re assuming. Altman’s ouster showed them the company was not as stable as they may have thought.
Then there is the recent exit of Andrej Karpathy, a founding member of OpenAI who rejoined the company only a year ago. In an internal note to staff, Karpathy insisted that he was not leaving for another job and that he was “very sad” to be leaving at all, according to a person who viewed it. The person noted that Karpathy simply was not enjoying the work he was tasked with at OpenAI, leading a team developing an AI assistant. Yet, the person wondered why the company wouldn’t have wanted to keep someone of his stature within the AI community — “Why not let him do whatever he wanted?”
The unsatisfying exit of Karpathy is in addition to the ongoing invisibility of Ilya Sutskever, OpenAI’s respected cofounder and longtime chief scientist who played a big role in Altman’s ouster. There is continued silence surrounding his future at the company, if any, both of the people familiar said.
OpenAI has yet to officially address Sutskever’s employment status. While he is no longer on the OpenAI board (nor is Altman), his name remains in company systems like Slack, one of the people familiar said. A lawyer for Sutskever did not respond to a request for comment.
With more questions than answers, most employees are left to observe and infer. As for what the internal investigation into Altman will result in, employees aren’t expecting any huge revelations, in part because Altman’s mood and demeanor have improved in recent weeks, one of the people said. The weight of “the coup” still appeared to “hang over” Altman at the start of this year, the person noted. Now, he “seems a lot happier.”
With Sutskever still out of the picture, other leaders at OpenAI are starting to appear front and center in his place. Greg Brockman now more frequently addresses workers and answers questions, both of the people familiar said. As does Jakub Pachoki, who heads research.
“Sam is moving on,” one of the people said. Everyone else is trying to, as well.
Are you an OpenAI employee or someone with a tip or insight to share? Contact Kali Hays at [email protected] or on secure messaging app Signal at 949-280-0267. Reach out using a non-work device.
