Chinese cybersecurity firm 360 Security Technology has unveiled two artificial intelligence-powered cybersecurity tools that it described as China’s answer to Anthropic’s advanced Mythos model even as the US-based AI firm has accused Alibaba of “illicitly” accessing its Claude AI model.According to a Reutersreport, speaking at the ISC.AI 2026 cybersecurity conference in Beijing, 360 founder Zhou Hongyi introduced “Tulongfeng”, an AI system designed to automatically discover software vulnerabilities, and “Yitianzhen”, a platform built to automate cyber defence and incident response.Zhou described Tulongfeng as “China’s version of Mythos”, arguing that vulnerability-discovery AI has become a strategic national capability.“This kind of powerful weapon that can change the landscape of cyber offence and defence cannot be held only by others,” Zhou said, according to a transcript published by the company.He said China risked “one-way transparency” if US organisations could deploy Mythos-like systems to identify vulnerabilities in software and critical infrastructure while Chinese companies lacked similar capabilities.According to 360, Tulongfeng has identified 3,432 software vulnerabilities, including 105 confirmed by Chinese authorities.
Anthropic accuses Alibaba of AI model extraction
Meanwhile, the launch comes as Anthropic has accused Alibaba of conducting what it described as the largest known effort to extract capabilities from its Claude AI models.In a letter seen by Reutersand sent to US Senators Tim Scott and Elizabeth Warren, Anthropic alleged that operators affiliated with Alibaba and its AI unit Qwen carried out a large-scale “distillation” campaign between April 22 and June 5, 2026.Anthropic said the operation generated more than 28.8 million interactions with Claude using nearly 25,000 fraudulent accounts. The company alleged the effort was aimed at training less capable AI models using Claude’s outputs to accelerate China’s progress toward advanced systems such as Mythos Preview.The allegations follow similar claims made by Anthropic earlier this year against Chinese AI companies, including DeepSeek, Moonshot AI and MiniMax. The company has previously warned that such distillation efforts could accelerate the development of competing frontier AI models.The developments come amid growing US-China tensions over AI leadership and technology controls. The US government has tightened export restrictions on advanced AI chips since 2022, while Anthropic’s Mythos model has drawn international attention over its ability to identify software vulnerabilities. The US has since restricted exports of a public version of Mythos on national security grounds.
