Fired OpenAI safety researchers say they were never told what leaked
Original title: Three fired OpenAI safety researchers say they were never told what they actually leaked
OpenAI says a leak probe, not safety advocacy, drove the firings; the three deny sharing model details.
Tomek Korbak, Mikita Balesni, and Jasmine Wang told the Wall Street Journal they were never shown in writing what they allegedly leaked before OpenAI fired them. Investigators suspected they shared internal architecture details with Ryan Greenblatt and found Signal chats with disappearing messages on their phones. OpenAI says the dismissals were unrelated to safety concerns and that Signal was not the reason. The researchers tie the dispute to outside access to chain-of-thought traces, including analysis around model Astra.
Unconfirmed: The researchers deny leaking information, while OpenAI says the firings followed a leak investigation and were unrelated to safety advocacy.
- All three were fired the same day, a week after investigators searched their phones.
- Balesni says investigators named Greenblatt as the suspected source of a leak.
- Greenblatt denies passing confidential information to journalists.
- Wang says she accidentally opened one executive email and reported it immediately.
- OpenAI redacted nearly half of a METR and Redwood Research on-site audit.
- Vendors
- OpenAI
- Organizations
- METRRedwood Research
- AI models
- Astra
From the original article
3 min read at the-decoder.com
The three OpenAI safety researchers who were fired, Tomek Korbak, Mikita Balesni, and Jasmine Wang, are publicly pushing back against allegations from their former employer. They deny sharing confidential information and see a connection to their advocacy for AI systems that are easier to scrutinize. The researchers say OpenAI never told them in writing which information they allegedly mishandled.
By Manuel Uth · the-decoder.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.