Anthropic Urges Congress to Crack Down on AI Distillation By Chinese Rivals

Anthropic urged Congress on May 14, 2024, to implement regulations that would prevent Chinese companies from using "distillation" to train AI models on stolen data. The AI safety company stated that Alibaba-affiliated operators created approximately 25,000 fake accounts to generate 28.8 million interactions with its Claude chatbot. This practice, known as distillation, allows competitors to train their own AI models by mimicking the outputs of more advanced, proprietary models without incurring the significant research and development costs. Anthropic's CEO, Dario Amodei, highlighted this issue in a letter to lawmakers, emphasizing the need for legislative action to protect intellectual property and ensure fair competition in the AI industry. The company believes that without such measures, the AI landscape could be dominated by entities that bypass ethical development and data acquisition processes. This concern underscores a broader debate about AI governance and the potential for malicious actors to exploit AI advancements.