Chinese firm Qihoo 360 has made a bold claim regarding the development of an AI system called “Tolongfeng.” 

According to the CEO, Zhou Hongyi, the technology is a "Chinese version" of Anthropic’s powerful "Mythos" AI, characterizing both as "cyber nuclear weapons" and strategic deterrents in the digital age. 

The system is reportedly capable of discovering software vulnerabilities. Zhou claims Tulongfeng has already identified over 3,000 vulnerabilities, some labeled "high risk."

Zhou also noted that while this technology may not match Mythos in raw power but in combination with Qihoo’s existing technologies, it could create an unprecedented hacking capability rivaling Anthropic’s system. 

He called the development of AI-driven cyber systems a new nuclear arms race, stating, "While nuclear weapons historically served as the primary strategic deterrent, the future of national security will likely be defined by vulnerability discovery capabilities."

Qihoo 360 is currently on US blacklist due to its alleged ties with the Chinese military and intelligence agencies. However, the company denied all these accusations. 

The claims came after the Five Eye security alliance comprising the US, UK , New Zealand, Canada and Australia issued a stark warning that adversaries are just months away from using AI to launch “overwhelming cyberattacks.” 

Alan Woodward, a cyber security expert at the University of Surrey, said with certain clarity that China would develop such systems. 

"It might not be as good as Mythos, but what it does show is these things are going to come out anyway," he said.