US Highlights Alleged Chinese AI IP Theft
- US cites 'industrial-scale' theft of AI tech by China
- China alleges US claims are unsubstantiated slander
- Distillation attacks raise concerns over AI competition
The United States has announced plans to address what it describes as China’s "industrial-scale theft" of intellectual property linked to American artificial intelligence labs. Allegations have surfaced following the launch of DeepSeek, a Chinese AI model that reportedly utilized outputs from existing models such as those from OpenAI.
Several companies have accused rivals, particularly from China, of employing techniques known as distillation to replicate proprietary AI systems. Notably, Google reported attempts to clone its Gemini chatbot while Anthropic pointed fingers at Chinese firms for engaging in similar practices to generate substantial amounts of data fraudulently.
US officials are alarmed by these tactics, cautioning that they could enable China to rapidly advance in the AI sector. In an internal memo, Michael Kratsios, director of the White House Office of Science and Technology Policy, emphasized the serious threat posed by these industrial-scale operations.