OpenAI Fires Two Researchers Over Information Leaks

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.

AI researcher Leopold Aschenbrenner. Image credit: Leopold Aschenbrenner

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.

AI researcher Pavel Izmailov. Image credit: Pavel Izmailov

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.

Matthew Broersma

Matt Broersma is a long standing tech freelance, who has worked for Ziff-Davis, ZDnet and other leading publications

Recent Posts

EU Widens Investigations Into Chinese Imports, Subsidies

After the United States imposes 100 percent tariffs on certain Chinese goods, Europe widens its…

23 hours ago

Reddit Deal With OpenAI Gives ChatGPT Access To Content

OpenAI strikes deal with Reddit to train its AI tech on user posts and give…

24 hours ago

Microsoft Invests 4 Billion Euros In France For AI, Cloud

Global spending spree from Microsoft continues, with huge investment for new data centre to drive…

1 day ago

Toshiba Axes 4,000 Staff In Post-Delisting Restructuring Operation

Workforce blow. Newly privatised Toshiba has embarked on a 'revitalisation plan' that will entail the…

2 days ago

European Union Opens Child Safety Probe Into Meta

European Commission opens an official child safety investigation into Facebook and Instagram-owner Meta Platforms

2 days ago