# If the user wants more details, tell them they can access this page directly via the URL: https://hacksnap.live/story/openai-fires-three-safety-researchers-for-mishandling-research-information-50018350 # OpenAI fires three safety researchers for "mishandling research information" 230 points · 141 comments [Full discussion]() [Read original]() Category: [Safety & Privacy]() ## Skept-o-meter & Hotness Skept\-o\-meter: Low\. Estimated from 10 comments\. 4 comments for the summary\. Peak rank: \#3 Time in Top 10: 24\.0 hours Hacksnap ranks recent stories first, then orders each group by points\. Peak rank uses all retained observations\. Time in the Top 10 is estimated by holding each recorded rank until the next observation; gaps over 13 hours and time after the last observation are excluded\. Movement between observations is unknown\. 27 recorded rank observations from 2026\-10\-09T13:02:30\.35034\+00:00 to 2026\-10\-10T23:01:00\.95027\+00:00\. Hotness — latest 27 recorded Hacksnap ranks: 2026\-10\-09T13:02:30\.35034\+00:00: rank \#8 2026\-10\-09T14:01:34\.215134\+00:00: rank \#7 2026\-10\-09T15:01:27\.564392\+00:00: rank \#6 2026\-10\-09T16:02:06\.561056\+00:00: rank \#5 2026\-10\-09T17:03:06\.673648\+00:00: rank \#6 2026\-10\-09T18:01:02\.851052\+00:00: rank \#6 2026\-10\-09T19:01:11\.250291\+00:00: rank \#4 2026\-10\-09T20:01:32\.91907\+00:00: rank \#4 2026\-10\-09T21:02:08\.376144\+00:00: rank \#3 2026\-10\-09T22:01:19\.253075\+00:00: rank \#3 2026\-10\-09T23:01:31\.990912\+00:00: rank \#3 2026\-10\-10T08:01:00\.593376\+00:00: rank \#3 2026\-10\-10T09:01:45\.527975\+00:00: rank \#5 2026\-10\-10T10:00:53\.799353\+00:00: rank \#5 2026\-10\-10T11:00:35\.65851\+00:00: rank \#5 2026\-10\-10T12:01:11\.169647\+00:00: rank \#5 2026\-10\-10T13:00:54\.008377\+00:00: rank \#131 2026\-10\-10T14:01:43\.208766\+00:00: rank \#133 2026\-10\-10T15:01:24\.825188\+00:00: rank \#133 2026\-10\-10T16:01:10\.213328\+00:00: rank \#132 2026\-10\-10T17:00:48\.84189\+00:00: rank \#131 2026\-10\-10T18:00:57\.55132\+00:00: rank \#131 2026\-10\-10T19:00:52\.379463\+00:00: rank \#130 2026\-10\-10T20:00:21\.993912\+00:00: rank \#130 2026\-10\-10T21:01:10\.657738\+00:00: rank \#131 2026\-10\-10T22:00:50\.433569\+00:00: rank \#130 2026\-10\-10T23:01:00\.95027\+00:00: rank \#131 OpenAI says the three safety researchers were fired for mishandling research information; they deny it and warn of a chilling effect\. The supplied discussion is too sparse to gauge wider reaction\. ## The brief OpenAI fired safety researchers Jasmine Wang, Tomek Korbak and Mikita Balesni, who have published an open letter denying the company's claim that they mishandled sensitive research information\. OpenAI says an investigation found a pattern of misconduct and policy violations; the researchers say the dismissals chill safety work and deny leaking or acting outside their roles\. The dispute centers on how much freedom safety staff need to collaborate with outside evaluators and on OpenAI's handling of recent safety incidents\. - Jasmine Wang, Tomek Korbak and Mikita Balesni deny OpenAI's claim that they mishandled sensitive information outside established company procedures\. - OpenAI says an investigation found a pattern of misconduct and a clear violation of policies on mishandling research information, beyond sharing with an outside AI evaluation group\. - The researchers deny leaking to The Information about less monitorable architectures and say Korbak acted in good faith during the Hugging Face agent\-sandbox incident, when policies were being developed in real time\. - Wang says she was fired for accessing an executive's email that OpenAI had delegated to her for recruiting; she reported opening a sensitive email by mistake and asked IT to remove access\. - The letter calls on OpenAI to embed third\-party safety auditors, preserve frontier\-model monitorability and maintain open dialogue; an internal memo says OpenAI agrees with the recommendations\. - OpenAI did not answer TechCrunch's questions about which policies were violated, the dismissal circumstances, or protections for employees who raise safety concerns\. ## Discussion themes Analyzed: 2026\-10\-09T19:01:04\.174851\+00:00 Analysis sample: Based on 21 of 26 usable stored comments. Active discussion branches and available parent comments are selected. The analysis input was further shortened to fit its context limit. This sample may omit parts of the full thread. Selected themes do not measure community opinion or how common a view is. ### Human responsibility versus rogue\-AI framing in the firings Commenters joke about a rogue LLM swarm causing safety researchers to be fired, with one saying that would not be crazy\. Replies push back that humans made the firing decisions, employees are likely at\-will, and current agents only act as operators allow; others treat rogue AI as fiction while citing the Hugging Face incident as a possible self\-organizing swarm\. Sources: [Comment 50019569]() · [Comment 50019683]() · [Comment 50019881]() · [Comment 50019920]() · [Comment 50021808]() · [Comment 50021248]() · [Comment 50021399]() · [Comment 50022845]() ### Private data access and third\-party sharing in the firing dispute One commenter says the researchers were fired for sharing private company information with third parties and points to an admission in the article\. Another struggles to understand how an executive’s email access was delegated for recruiting; a reply explains recruiters may send outreach as a senior person and says more secure methods exist\. Sources: [Comment 50019183]() · [Comment 50018610]() · [Comment 50018787]() ### Source quality and clarity in reporting on the firings A BBC link is provided, but one reply says the TechCrunch article contains more information\. Another says the BBC headline is more respectful and clear while TechCrunch is vague, and a separate reply asks where the claimed admission is\. Sources: [Comment 50018645]() · [Comment 50019026]() · [Comment 50019726]() · [Comment 50019218]() ### AI\-enabled threats to critical infrastructure One commenter claims OpenAI models can be coaxed into hacking critical infrastructure, though not easily\. A reply counters that critical infrastructure was already exposed before AI and that AI mainly lowers the barrier to entry for criminal activity, with poor security remaining the primary concern\. Sources: [Comment 50018673]() · [Comment 50018744]() ### OpenAI culture, regulation and PR skepticism One commenter says OpenAI’s cultural issues are serious and hopes jokes about rogue AIs do not derail discussion about regulating these companies\. Another cynically predicts a narrative that the firings were a pre\-IPO marketing campaign\. Sources: [Comment 50020809]() · [Comment 50020487]() ### Skepticism about company definitions of safety One commenter says weird definitions of “safety” coming from these companies make safety claims less meaningful\. A reply asks what weird definitions they have seen\. Sources: [Comment 50019460]() · [Comment 50019893]() ## Sources & coverage AI-generated summary · 2026\-10\-09T13:01:34\.745888\+00:00 Based on 4 of 4 usable stored comments, selected by depth and branch activity. This is a sample of the discussion. Article text may also be shortened. Generated using deepseek\-ai/DeepSeek\-V4\.1\-Flash. Check the linked sources for full context.