Access to an online forum for OpenAI employees provided input into the design of AI technology.
The hack happened early last year, according to sources at the New York Times, at a forum where employees discuss the latest technology. The Times clarifies that the hackers did not have access to the systems where the AI applications are stored and built.
National Security Concerns
After the incident was observed, OpenAI executives notified employees at a meeting in April 2023. The board of directors was also notified. Management decided to keep the incident largely secret and not publicize it because it did not involve customer or partner information. It was also not considered a threat to national security because the hackers acted privately. Regulatory authorities were not notified.
The fact that the hackers acted independently and had no ties to foreign governments is considered necessary for sharing information with authorities. But the news has some OpenAI employees worried that countries like China could steal the AI technology, which could jeopardize U.S. national security. At the same time, the incident has raised questions about how seriously OpenAI takes security.
Reuters and other news agencies have asked OpenAI for a response but have not yet received one.
Tip: CriticGPT finds errors in AI model output
