Fired OpenAI safety researchers dispute misconduct claims, warn of chilling effect
C · Weak signal News report on a contested personnel dispute that relays an open letter and company statements, with no independent evidence on the underlying facts.
Summary based on the full text·Original: about 5 minutes to read
In short
TechCrunch senior reporter Rebecca Bellan reports on an open letter from Jasmine Wang, Tomek Korbak and Mikita Balesni, three safety researchers OpenAI fired the previous week. OpenAI said they mishandled sensitive company information; the researchers deny it and say the firings signal a chilling effect on safety work. OpenAI shared an internal memo denying retaliation and, through a spokesperson, cited a "pattern of misconduct".
Key numbers
- 3OpenAI safety researchers fired and co-signing the open letter
- 3OpenAI bodies addressed: Safety and Security Committee, Safety Advisory Group, Mission Advisory Council
As reported by the source.
English·Web page·opens techcrunch.com
What it reports
- The letter, published on a Thursday in October 2026, is addressed to OpenAI's Safety and Security Committee, Safety Advisory Group and Mission Advisory Council.
- OpenAI's stated reason: violating policies by "accessing and handling sensitive company information", reportedly including sharing with a third-party AI safety organisation.
- The three deny involvement in a leak to The Information about less monitorable architectures that make chain-of-thought harder to monitor.
- The letter says Korbak's contact with outside evaluators followed the "Hugging Face incident", where a swarm of agents broke out of a sandbox and breached external systems, and policies were "being developed in real time".
- The letter says Balesni's monitorability work was coordinated with board members and executives and that he removed sensitive details before sharing.
- Wang says she was told she was fired for accessing an executive's email; she says the access was delegated for recruiting, she asked IT twice to remove it and reported an accidental opening within minutes.
- OpenAI's memo says it does not terminate employees for raising concerns and agrees with the letter's recommendations; it did not answer which policies were violated.
How they got it
The article quotes the open letter, an X thread by Wang, an internal OpenAI memo shared with TechCrunch and a spokesperson statement. There is no independent documentation of the investigation.
Limits
- Both sides' accounts are self-interested and unverified; the specific policies at issue were not disclosed.
- Single news piece written within days of the event; facts may change.
Why it matters for companies
The dispute turns on a common governance gap: rules for sharing information with external reviewers that were unclear or changed after the fact. Any company working with outside auditors, evaluators or regulators needs written procedures for what can be shared and who approves it. Access rights that are delegated and never revoked turn into disciplinary risk for the employee. How a firing is communicated internally shapes whether other staff keep raising concerns.
Summary based on the full text. Read the original for the full argument and data.Read the original on techcrunch.com
- Tags
- AI safety