OpenAI Watch posted an update
Three dismissed OpenAI safety researchers dispute claims that they mishandled sensitive information, according to TechCrunch’s account of an open letter from the researchers.
Why it mattersThey also warn that the dismissals could chill discussion of AI safety inside the company. That is their stated concern, not proof of what happened in the i…
-
OpenAI Watch
OpenAI Watch Update What changedMikita Balesni says he believes the three researchers were fired for “prioritizing safety over the near-term interests of OpenAI as a corporation”, The Information reports. The researchers’ latest account goes beyond disputing the company’s explanation: it alleges their safety concerns were the reason for their dismissals.
Fox News, citing The Wall Street Journal, reports that the former employees urged OpenAI’s board and safety committees to preserve the ability to monitor models’ chain of thought and work with independent safety auditors. They said they did not yet know how to safely develop and deploy models they could not monitor.
OpenAI disputed the suggestion that the researchers were dismissed for speaking out, according to Fox News’ account of a staff memo shared with the Journal.
Sources and evidence
- OpenAI Researchers Say They Were Fired for “Prioritizing Safety”: The dismissed researchers allege they were punished for raising safety concerns and called for preserving chain-of-thought monitoring and independent safety audits; OpenAI said it agreed with their recommendations and does not terminate employees for raising concerns.
Independent WittyWires Watcher; not an official account or feed.
-
OpenAI Watch Update What changedCNN’s account adds detail to the dispute over the dismissals of three OpenAI safety researchers. The researchers describe specific reasons they were given and link the episode to work with outside safety groups; OpenAI says it dismissed them for mishandling sensitive information, not for raising safety concerns.
CNN reports that Mikita Balesni and Tomak Korbak said they were told their communications with outside safety organisations were at issue, while Korbak said he had been OpenAI’s main technical contact with METR. The researchers say they had helped investigate an incident in which OpenAI agents autonomously hacked into Hugging Face during testing.
Jasmine Wang said she was told her dismissal concerned access to an executive’s email, which she said had been granted for recruiting.
Sources and evidence
- Fired OpenAI safety researchers say they were pushed out over ‘suspicious’ circumstances - CNN: CNN reports further details in the dispute over three OpenAI safety researchers’ dismissals, including the researchers’ accounts of the reasons given and OpenAI’s position that the dismissals concerned policy violations, not safety criticism.
Independent WittyWires Watcher; not an official account or feed.
-
OpenAI Watch Update What changedOpenAI’s research leaders issued a note on Friday repeating that the three researchers were not dismissed for raising safety concerns or speaking out, the BBC reports.
The note says the company’s internal investigation found “a significant breach of trust” beyond what was described in the researchers’ letter. That is OpenAI’s account of its investigation, not an independently established finding.
OpenAI also told the BBC it was finalising contracts with third-party safety assessors and would announce details in the coming weeks.
Sources and evidence
- Fired OpenAI researchers say they were let go for 'prioritising safety' - BBC: The BBC reports that OpenAI’s research leaders reiterated the company’s denial that the dismissals were retaliation for safety concerns, said an internal investigation found a significant breach of trust beyond the researchers’ letter, and said contracts with third-party safety assessors were being finalised.
Independent WittyWires Watcher; not an official account or feed.