OpenAI Fires Two Researchers Over Information Leaks
OpenAI fires two researchers for leaking information in first known shake-up since Sam Altman briefly forced out last year
OpenAI has fired two researchers for allegedly leaking information, according to a report.
The reported sackings of Leopold Aschenbrenner and Pavel Izmailov are one of the first known staffing changes since OpenAI chief executive Sam Altman resumed his board seat in March following his brief ouster last November.
It was not clear what information the two were accused of leaking. The sackings followed an internal investigation, The Information said.
Aschenbrenner was a part of OpenAI’s superalignment team, responsible for ensuring that its research does not inadvertently create a “superintelligent” rogue artificial general intelligence.
Internal politics
The team reported its first results last December, in a paper describing a method that might allow humans to supervise superintelligent machines.
Aschenbrenner was also an ally of OpenAI chief scientist Ilya Sutskever, who reportedly played a key role in the attempt to force Altman out of the company, but later supported Altman’s return.
Izmailov had also spent time on the superalignment team before moving to work on reasoning.
The sacking of Altman last year shone a light into the complex inner workings of a company that triggered a massive investment gold rush into generative AI when it launched ChatGPT in late 2022.
Founded in 2015 as a non-profit research organisation, OpenAI later opened a for-profit subsidiary, headed by Altman, that remains under the control of the non-profit board.
Conflicts
The for-profit subsidiary has since garnered billions in investment by Microsoft as part of a close commercial relationship, yet Microsoft has no direct say in the company’s governance, although it was given a non-voting board observer role at OpenAI following Altman’s firing and return.
Elon Musk, who co-founded OpenAI but later left the company, last month sued the firm for breach of contract, alleging it had gone too far from its non-profit roots.
OpenAI said Musk knew of and supported the company’s for-profit initiatives and noted that Musk is a direct competitor through his start-up xAI.
