OpenAI defended its firing of three safety researchers who raised concerns about AI societal threats, saying the trio committed a “significant breach of trust.”
The AI giant said Friday that its decision to part ways with the eggheads wasn’t because they had raised safety concerns, but because their actions went “beyond what’s outlined in the letter they published and we stand by the decision to not continue their employment.”
The row comes as AI safety concerns have reached a fever pitch following a recent string of hacking incidents by leading companies including Sam Altman’s OpenAI and Dario Amodei’s Anthropic.
OpenAI’s statement came in response to accusations from the three researchers saying they’d been “fired for prioritizing safety over the near-term interests of OpenAI as a corporation.”
“We have become concerned that internal and external communications around our firing have made our former colleagues afraid to speak and operate in way that, until last week, were an integral part of working at OpenAI,” reads a letter from the researchers that was posted online Thursday.
Mikita Balesni, one of the fired researchers, added on X: “In the exit call, I was told OpenAI no longer trusts me because I was speaking too much to third party safety organizations, implying I leaked company IP.
“I never shared company IP. The work I was doing was coordinated with my reporting line, research leadership, and the board,” he added.
Balesni continued that he was concerned OpenAI would use the firing as an excuse to cut off the company’s relationship with AI watchdog Model Evaluation and Threat Research, or METR. He said he worried that OpenAI might not “give the most competent safety auditors continuous employee-level access that @sama promised,” tagging OpenAI CEO Sam Altman’s X account.
METR is one of two prominent AI safety watchdogs — along with Redwood Research – that collaborated to release a bombshell report last month detailing how a serious hack of AI company Hugging Face occurred.
The Post has detailed the cozy relationship those companies have with AI titans and their tight ties to the cult-like “Effective Altruism movement,” which has counted disgraced crypto fraudster Sam Bankman-Fried among its devotees.
The back-and-forth was kicked off last week after OpenAI canned the three researchers – Jasmine Wang, Tomek Korbak and Balesni – after they allegedly shared confidential information with a third-party AI safety group. It wasn’t immediately clear which AI safety group received the confidential information.
AI safety concerns escalated in September after researcher Jacob Coxon quit his role at Anthropic, warning that advanced AI “could kill us all by the end of the decade.”
Coxon, who previously worked at OpenAI, said in an thread on X that “neither company is acting responsibly.”
After Coxon issued his public warning, Amodei issued a call to “pace the frontier” with an industrywide slowdown.
