San Francisco: OpenAI, a U.S.-based artificial intelligence firm, announced that it has banned several user accounts potentially linked to North Korea. This action was taken in response to a deceptive employment scheme aimed at exploiting OpenAI's AI models for malicious purposes.
According to Yonhap News Agency, OpenAI disclosed in an updated report that the actors involved were crafting personal documentation for fictitious job applicants. This included creating resumes, online job profiles, and cover letters as part of a scheme attributed to North Korea. The intention was to deploy IT workers overseas to generate hard currency for the Pyongyang regime.
The report, titled "Disrupting malicious uses of our models," noted that the activities observed aligned with tactics identified by Microsoft and Google. While OpenAI could not determine the exact locations or nationalities of the actors, the characteristics of the activities mirrored publicly reported North Korean efforts to fraudulently obtain positions at Western companies to support its financial network.
The actors also created support personas to provide reference checks and refer the fictitious applicants for employment opportunities. Additionally, they crafted social media posts to recruit real individuals willing to participate in the scheme by hosting laptops or lending their identities to help the applicants pass background checks.
OpenAI's AI models were also reportedly used to generate plausible responses to technical and behavioral interview questions. This report comes amid efforts by Seoul and Washington to curb Pyongyang's use of IT workers overseas through deceptive practices, as such activities are believed to fund the North Korean regime's nuclear and ballistic missile programs.