Anthropic accuses Alibaba of 'illicitly' accessing its AI models

US-based AI firm Anthropic has accused Chinese firm Alibaba of illicitly accessing its Claude AI models' capabilities. Anthropic, in a letter, said Alibaba and its AI lab carried out 28.8 million exchanges with its models using 25,000 fraudulent accounts. Anthropic said Alibaba carried out "distillation attacks", which use outputs from a stronger AI model to train a weaker one.