On July 4, the New York Times reported that OpenAI experienced a security breach in 2023, but did not disclose the incident to external parties. It is reported that OpenAI executives disclosed the incident internally at a meeting in April 2023, but did not publicly disclose it because the attackers did not access information related to customers or partners. In addition, executives did not believe that the incident posed a national security threat because they believed the attackers were individuals not affiliated with foreign governments. They did not report the incident to the Federal Bureau of Investigation or other law enforcement agencies.
It is reported that the attackers accessed OpenAI's internal messaging system and stole detailed information about the company's artificial intelligence technology design from employee conversations on online forums. They did not access OpenAI's "accommodation and construction of artificial intelligence" system or code.