US AI Firm Claims Chinese Labs Illegally Use Its Technology

Want the full story?
Read the complete article at KVIA
AI Summary
Anthropic, a US-based AI company, has accused three Chinese AI labs—DeepSeek, Minimax, and Moonshot AI—of illegally extracting features from its Claude model, raising significant national security concerns. The allegations include the creation of over 24,000 fraudulent accounts to train their models using 16 million interactions with Claude, a practice known as distillation, which is typically banned by leading AI providers. This follows similar claims by OpenAI regarding DeepSeek's alleged misuse of its ChatGPT models. Anthropic warns that these unregulated models could pose risks related to cybercrime and authoritarian surveillance, emphasizing the need for strict US export controls to maintain technological leadership.
Key Details: • Anthropic claims over 24,000 fake accounts were created to exploit its Claude model. • The allegations highlight concerns over national security and the potential misuse of AI technologies. • OpenAI has also raised similar concerns about DeepSeek's practices earlier this month. • Anthropic advocates for US export controls to prevent unauthorized access to advanced AI capabilities.