Michaela
A recent report has revealed that North Korea has been developing artificial intelligence capabilities since the late 1990s, partially using NVIDIA graphics cards restricted by export bans. Experts warn that this technology could enhance Pyongyang’s AI-driven operations, including cybercrime, cryptocurrency theft, money laundering, and the creation of fake identities, potentially intensifying the country’s digital threats.
Key Points
- North Korea has been developing AI since the late 1990s, reportedly using NVIDIA graphics cards banned by export restrictions. This technology could significantly boost its cybercrime and crypto theft operations.
- Pyongyang's AI capabilities, including deepfake creation and data optimization, are applied to cyber operations like crypto theft and money laundering. These efforts fund its nuclear and missile programs despite international sanctions.
- The report highlights the growing difficulty in curbing North Korea's technological ambitions. Restricted hardware, like NVIDIA GPUs, continues to surface in its research pipelines despite tightened U.S. export rules and bans.
According to local outlets, the report, titled Analysis of North Korea’s AI Status and Policy Considerations, found that the country’s AI research has focused on pattern recognition, voice processing, and data optimization. Since the 2010s, North Korea has bolstered its capabilities by expanding research institutions and developing domestically designed algorithms.
Notably, certain studies employed NVIDIA’s GeForce RTX 2700 graphics cards, which the U.S. Department of the Treasury’s Office of Foreign Assets Control (OFAC) fully banned from export or re-export to North Korea.
These AI capabilities could be applied across North Korea’s cyber operations, including deepfake creation, evading detection, crypto theft, and data optimization. In recent years, the regime has focused resources on these technologies to generate revenue for its nuclear and missile programs despite international sanctions. Research into multi-person tracking, which connects identities across separate video footage, could evolve into real-time automated surveillance when integrated with CCTV and drone systems.
Related: Disney to Invest $1B in OpenAI for AI-Generated Videos Featuring Iconic Characters
In April, reports suggested the Trump administration had begun reviewing a Biden-era policy limiting global access to U.S.-manufactured AI chips, including a possible move to scrap the existing tiered framework that governs how many advanced semiconductors individual countries are permitted to purchase.
The U.S. Department of Commerce released its Framework for Artificial Intelligence Diffusion in January, just days before the end of the Biden administration. The policy aims to curb the international spread of high-performance AI chips by assigning access levels according to national security risk.
Related: Time Names AI Innovators ‘Architects of AI’ as 2025 Person of the Year
The framework also places restrictions on certain AI model weights to keep the most advanced computing resources within the United States and allied nations, while limiting their availability to countries including China.
The report’s findings spotlight the growing difficulty of curbing North Korea’s technological ambitions, as restricted hardware like NVIDIA GPUs continues to surface in opaque research pipelines despite tightened export rules.
MICHAELA
Michaela is a news writer focused on cryptocurrency and blockchain topics. She prioritizes rigorous research and accuracy to uncover interesting angles and ensure engaging reporting. A lifelong book lover, she applies her passion for reading to deeply explore the constantly evolving crypto world.
