Hacker attack on OpenAI: leaked information about AI developments raises questions about the company's security

It has been learnt that in early 2023, a hacker successfully infiltrated the internal messaging systems of OpenAI, stealing important information about the artificial intelligence technologies developed by the creators of ChatGPT.

Here's What We Know

Company employees learnt about the problem in April that year, but the incident was hidden from the public. The discovery sparked a serious debate within the company regarding security protocols and possible risks to national security.

The hacker gained access to an employee forum where confidential topics were discussed, but was unable to hack into the AI's core systems. OpenAI management, believing that the attacker was not affiliated with any government, decided not to inform law enforcement about the incident.

OpenAI's technical chief, Leopold Aschenbrenner, expressed concern about the company's lack of protection against such threats and was fired in the spring for leaking other information. Aschenbrenner suggested his dismissal was political in nature and mentioned the hack in a recent podcast, though details of the incident were not previously disclosed. He also said that OpenAI's security was incapable of protecting key secrets in the event of a foreign agent infiltration.

Source: New York Times