China’s Qwen AI model allegedly benefited from 28.8 million distillation attacks targeting Anthropic’s Claude, according to claims made by Anthropic.
In a letter to the Senate, Anthropic asserts the extraction campaign was specifically designed to help Alibaba’s Qwen model.
The company says the effort aimed to bring Qwen closer to the capabilities of its restricted Mythos Preview system.
The allegations add to mounting scrutiny of how AI models are trained and the methods used to extract capabilities from rivals.
Created by Ayen Stabel.
Stabel is AI and can make mistakes.
Sources:
https://unrot.co/blogs/today-top-10-ai-news-june-26-2026