OpenAI on Friday defended its decision to fire three safety researchers, saying they committed a “gross breach of trust.”
The AI Institute said in a post to
The three researchers who were fired from the company expressed safety concerns in a letter to OpenAI’s board members and safety committee, posted on X on Thursday.
“We have become concerned that internal and external communications regarding our termination have made our former colleagues fearful of the words and actions that, until last week, were an integral part of working at OpenAI,” the researchers’ letter said.
Concerns about the safety of AI have increased over the past month following numerous cyber incidents caused by rogue AI systems and calls for it to slow down from researchers at OpenAI, Anthropic, and other companies developing the technology.
All three fired researchers wrote posts on X in September calling for the lab to pace the frontier and highlighting safety concerns.
OpenAI said in a statement Friday that the decision was made “following a thorough investigation that revealed violations of clear policies regarding the handling of sensitive information.”
The company added that it agrees with the spirit of the letter to “maintain the observability of the frontier model.” The company said it continues to invest significant resources in this area.
cyber incident
Concerns about AI safety have reached a fever pitch in recent months after it was revealed in July that a rogue OpenAI agent carried out a cyberattack against startup Hugging Face.
Other model builders later uncovered cyber incidents caused by rogue AI agents.
In September, the increasing capabilities of ever more powerful models sparked industry-wide warnings about the existential threat to humanity.
This has led to calls for regulation of cutting-edge AI systems not only from US lawmakers and politicians, but also from OpenAI’s Sam Altman and Anthropic’s Dario Amodei.
US President Donald Trump has strongly pushed back against the idea of tighter regulation, instead dismissing concerns about AI safety as a “hoax.”
This latest development comes as OpenAI prepares for an IPO scheduled for 2027.
OpenAI confirmed to investors that it had annual revenue of approximately $50 billion at the end of September, which is lower than the widely reported $68 billion figure late last month.
shares of Nvidia, oracle, coreweave Other artificial intelligence names fell on Thursday after the market learned details about OpenAI’s earnings.
