A Hacker Stole OpenAI Secrets, Raising Fears That China Could, Too

In early 2023, a hacker infiltrated OpenAI's internal messaging systems and accessed discussions about the company's A.I. technologies. While the breach did not extend to systems where A.I. models were being developed, it raised security concerns among employees about potential espionage by foreign adversaries. Despite these concerns, OpenAI's executives chose not to disclose the incident publicly, believing the hacker to be an individual without foreign government ties. A subsequent internal memo questioned the adequacy of OpenAI’s security measures, highlighting internal divisions over the handling and risk of A.I. technology.

Opens on The New York Times