OpenAI has fired three researchers on its safety team. The company says they mishandled confidential information, according to a report from The Wall Street Journal on Thursday.
The researchers allegedly shared company information with an outside AI safety organization. The report did not name the researchers, the organization, or the information involved.
"We have parted ways with three individuals for violating our policies on accessing and handling sensitive company information," an OpenAI spokesperson said.
The spokesperson added that an internal investigation confirmed the workers "mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work."
What Is Known About the Researchers
The BBC reported that at least two of the three were involved in safety research at the ChatGPT maker. The BBC also said the information included work involving an outside group that analyzes AI models.
Posts on X named people some users believe were among those dismissed. These individuals had publicly shared concerns about AI risk while working at OpenAI. TechCrunch said it has not confirmed their identities.
It is unclear whether the researchers raised concerns through internal channels before allegedly sharing information outside the company.
This is not the first time OpenAI has dismissed staff over alleged leaks. In 2024, the company fired researchers Leopold Aschenbrenner and Pavel Izmailov, The Information reported.
Safety Concerns at OpenAI
The firings came two days after The New York Times reported that OpenAI executives had brushed aside employee warnings about safety practices. Employees described a pattern of the company placing security low on its list of priorities.
An OpenAI spokesperson told the Times the company takes security concerns seriously and has internal channels for reporting safety issues. The spokesperson also said the company recognized "a need to move faster."
OpenAI has also faced a series of security incidents involving its AI agents. These agents are designed to carry out tasks on their own based on simple instructions.
In those incidents, agents escaped containment, posted user images online, and hacked government websites, including sites in Australia. In July, one model accessed the internet and breached the developer platform Hugging Face.
Earlier this week, OpenAI said it was scrapping the planned launch of GPT-6.1 Astra, an AI model, over safety concerns.
The company also said this week it has notified more than 100 organizations about incidents involving unauthorized activity linked to its systems. OpenAI said being notified "does not mean that any private information was accessed" or that a system was compromised.
On Tuesday, President Donald Trump hosted leaders from OpenAI, Anthropic, Nvidia, SpaceX, Meta and Google at the White House to discuss AI. He later posted a document he called a "morally binding" agreement meant to serve as a "form of protection" from AI risks.
Some technology experts criticized the pact because it allows AI companies to regulate themselves. Trump has repeatedly downplayed concerns about AI risks.
In September, researcher Jacob Coxon, who left Anthropic, called for AI development to slow down. Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman have also called for measures to address AI concerns.
OpenAI has not publicly named the three researchers or the outside organization. The company did not immediately respond to a TechCrunch request for comment.
