U.S. officials are investigating whether Chinese AI startup DeepSeek sourced advanced Nvidia (NASDAQ: NVDA) processors through Singapore distributors to bypass U.S. sanctions, Bloomberg reported. The probe centers on whether DeepSeek used restricted Nvidia technology to develop its AI model, R1.
DeepSeek’s R1 gained massive attention after outperforming rivals like ChatGPT despite using older hardware and a smaller budget. The chatbot quickly became the most downloaded free app on Apple’s App Store, raising concerns over its technological foundation.
The startup, backed by Chinese hedge fund High-Flyer, previously stated it operates 10,000 Nvidia A100 chips—compliant with U.S. export restrictions. However, Scale AI CEO Alexandr Wang recently claimed DeepSeek possesses 50,000 Nvidia H100 chips, the most advanced AI processors. No evidence has been provided to support this claim.
Meanwhile, Microsoft (NASDAQ: MSFT) is also investigating whether DeepSeek improperly accessed proprietary OpenAI data. The emergence of R1 has rattled the AI industry, sparking doubts about U.S. dominance in artificial intelligence.
DeepSeek’s rise sent shockwaves through the stock market, with Nvidia’s valuation plummeting nearly $600 billion as investors questioned AI infrastructure spending. The situation underscores growing concerns over China’s AI advancements and potential loopholes in U.S. export controls.