OpenAI said Feb. 21 that it had identified and removed several North Korean-linked accounts that misused ChatGPT to create fake resumes and online profiles.
According to the company, the accounts used ChatGPT-generated materials to apply for jobs at Western companies. OpenAI did not disclose the number of accounts removed or when the deletions occurred.
North Korea has increasingly turned to generative AI to steal advanced Western technologies and fund its illicit nuclear weapons program. North Korean hackers have long targeted employees of cybersecurity and cryptocurrency firms through platforms like LinkedIn, seeking access to sensitive information and assets.
More recently, they have refined their tactics using generative AI. Hackers reportedly employed ChatGPT to craft fake recruiter profiles, build trust with targets over time, and eventually install spyware to siphon funds.
OpenAI also said it had removed accounts linked to China. Some of these accounts allegedly used ChatGPT to generate Spanish-language news articles critical of the United States. The articles were published under the names of Chinese companies on major Latin American news platforms.
Additionally, a Chinese security organization developed an AI-powered tool to monitor anti-China content on social media, OpenAI said. The company, which named the tool “Peer Review,” confirmed that ChatGPT was used during its debugging process.
OpenAI noted that this marked the first known discovery of an AI-based surveillance tool.
