Chinese e-commerce giant Alibaba has been accused by leading American AI-developer Anthropic of "illicitly" seeking to harvest the abilities of its Claude chatbot. Photo by Jessica Lee/EPA
June 25 (UPI) -- AI firm Anthropic sought the assistance of Congress to crack down on alleged clandestine efforts by Chinese e-commerce giant Alibaba to harness its Claude AI assistant to develop its own rival version.
The tech giant wrote the Senate Banking, Housing and Urban Affairs Committee accusing Alibaba of "brazenly" seeking to harvest its AI capabilities by using Claude's power to train a smaller AI chatbot it was building, according to CNBC, which said it had seen the June 10 letter.
Anthropic alleged that over a six-week period beginning April 22, Alibaba's AI lab and affiliated entities used around 25,000 fake accounts to initiate almost 29 million interactions with its models in the "largest known distillation attack on Anthropic to date."
The operation focused on capturing Claude's most prized abilities, such as processing larger and more complicated prompts and questions, as well as how it made decisions, said Anthropic.
