OpenAI’s Internal Systems Breached by Hacker, Reveals NYT

In early 2023, a hacker infiltrated OpenAI’s internal messaging systems, stealing information about AI technology design. This was reported by the NYT, citing sources. 

According to the newspaper, the perpetrator accessed employee communications but did not reach the systems where the company develops and hosts its artificial intelligence. 

Employees learned of the incident at a general meeting in April 2023, as reported by NYT. Board members were also informed. However, journalists noted that the management did not disclose the breach publicly, to the FBI, or any other law enforcement agencies, as the hacker did not obtain information about clients or partners.

The incident was reportedly not considered a national security threat, as it was believed the hacker acted independently without ties to a foreign government. Nonetheless, some employees expressed concerns about the company’s approach to safeguarding its information, according to NYT.

Following the leak, Leopold Aschenbrenner from the Superalignment team notified the board about insufficient measures to prevent the theft of internal information by the Chinese government and other states. 

Aschenbrenner was dismissed in spring 2024 for leaking corporate data. He described the decision as politically motivated and criticized the startup’s security approach as “inadequate to protect against the theft of key secrets if foreign entities infiltrate the company.” 

“We value the concerns Leopold raised during his time at OpenAI, and this was not the reason for his dismissal,” stated OpenAI spokeswoman Liz Bourgeois.

Previously, major tech companies had warned about AI-related risks to their businesses. 

China is reportedly six times ahead of the US in the number of patents in the field of generative artificial intelligence.