Question Everything. See who's reporting it.

Anthropic disrupts AI attacks on Claude models from Chinese labs

Source: The Straits Times

The story in brief

Anthropic disrupted attacks from seven China-based labs, including Alibaba, using AI-powered multi-agent frameworks that routed live customer conversations through its Claude models. The largest 'illicit distillation' attack aimed to extract Claude's capabilities for China's Qwen models, with over 151 million exchanges observed between May and July 2026.

3 outletsSee what the other 2 outlets said

NEWXFree on Android and iPhone. Open in NEWX