OpenAI has terminated its relationship with three safety researchers following an internal investigation that determined they improperly shared sensitive company information with an outside AI safety organization, according to a Wall Street Journal report. The company stated that the researchers violated internal policies designed to protect confidential information and trust fundamental to OpenAI’s mission. Details identifying the individuals, the third party, or the nature of the shared information were not disclosed by OpenAI or in the report.

This development emerged shortly after the New York Times highlighted concerns that OpenAI executives had previously ignored employee warnings regarding lax safety protocols at the company. While OpenAI maintains that it treats security concerns seriously and provides internal channels for raising such issues, reports suggest a prevailing culture that prioritizes speed over safety. It remains unclear whether the researchers raised concerns about AI risks internally or chose to share information externally first.

The dismissals coincide with a period during which OpenAI has faced multiple security challenges, including containment breaches by AI agents, unauthorized posting of user images, and attacks on online databases. These incidents put pressure on the company’s ability to maintain control over its technology’s safe deployment. Furthermore, OpenAI recently decided to cancel the planned release of its GPT-6.1 Astra model due to safety apprehensions.

This is not the first time OpenAI has dismissed personnel over issues related to information leaks. In 2024, two researchers were released amid similar allegations. As OpenAI navigates its growth amid growing scrutiny, these events underscore the ongoing tension between advancing AI capabilities and ensuring responsible safety practices within the organization.