North Korea uses ChatGPT to infiltrate Western companies

OpenAI has blocked accounts of the ChatGPT chatbot linked to a South Korean scam where North Koreans impersonate IT workers from other countries to get into large Western companies, Axios reports.

North Korean IT workers often pose as American citizens to get remote work at Western companies. In addition to funding the state, this scheme allows them to infect company systems, as was the case with KnowBe4, and to collect confidential information.

OpenAI was able to find that the accounts of North Korean users who were blocked were used to optimize each stage of the scam. ChatGPT was used to craft cover letters, code, set up VPNs, and even write scripts to keep work laptops active and visible online.

The scammers also used a chatbot to mass-build resumes based on specific job openings, skill templates, and created profiles of US citizens.

Another use of ChatGPT is to recruit American citizens to create laptop farms that were issued to North Korean workers by companies. Incidentally, in 2024, a similar scheme was uncovered in the US, involving a Ukrainian who helped fake identities.

In April, Google issued a warning that North Korean IT workers were also targeting European companies, posing as Ukrainians to get jobs there. In addition, North Korea has created a separate hacking unit that focuses on AI-powered hacks.