North Korean cybercriminals are incorporating AI into their tactics, aiming to steal advanced technologies and funds for the country's illicit nuclear weapons program. Previously targeting global defence, cybersecurity, and crypto company employees, the hackers are known for operations like the Bangladesh central bank heist and the WannaCry ransomware attack on the UK's National Health Service in 2017.
OpenAI and Microsoft confirmed that North Korean, Chinese, Russian, and Iranian hackers use their AI services for malicious cyber activities. Moreover, South Korea has identified North Korean hackers using generative AI to target security officials, presenting a new challenge. Despite North Korean hackers' previous limitations in conversational English or Korean, generative AI allows them to create credible-looking profiles on platforms like LinkedIn, enhancing their phishing and social engineering operations.
Sophisticated attacks involve detailed profiles on professional networking sites, such as LinkedIn, where hackers build relationships over weeks and months. Using generative AI tools enables North Korean hackers to pose as recruiters and engage targets in technical exercises that lead to the installation of spyware. Platforms like LinkedIn serve as effective hunting grounds for fake North Korean recruiters, but they also leverage other platforms like Facebook, WhatsApp, Telegram, and Discord.
AI services like ChatGPT could aid North Korean hackers in developing more sophisticated forms of malicious software or malware. While safeguards exist to prevent misuse, individuals have found ways to circumvent them. North Korea, with access to Chinese AI services, has invested heavily in developing its cyber capabilities, using funds raised from criminal cyber operations to support its ballistic missile and nuclear programs.
Furthermore, North Korean researchers have published numerous AI-related studies, indicating the country's interest in AI applications. Academic papers from 2022 explore machine learning methods like reinforcement learning in war gaming simulations and the safe operation of large nuclear reactors using different machine learning techniques. The sophistication of North Korean AI systems is in its early stages, with the possibility that the country is intentionally concealing its capabilities.