Chinese cybersecurity researcher Yuki Chen, a five-time leader of Microsoft's security researcher rankings and two-time winner of the "Oscars of hacking," has joined Chinese AI firm Z.ai.
Chen, also known online as Guhe, announced his move to the Beijing-based company on social media last week. "New journey, new mission!" he wrote, as quoted by the South China Morning Post. "Thrilled and grateful for the opportunity to help build a safer cyberspace with the power of frontier AI."
Z.ai said Chen and his team recently identified 23 security vulnerabilities in Microsoft software using its GLM artificial intelligence models, with Microsoft acknowledging the discoveries.
"[Chen’s] decades of vulnerability research have helped, and will keep helping, the community build a safer cyberspace with GLM," the company said but it did not disclose how much of the research was conducted by AI or how extensively human researchers were involved.
Chen previously worked as a security researcher at Trend Micro and served as chief technology officer of Chinese cybersecurity firm Cyber Kunlun, according to Chinese technology publication Zhidx.
In 2015, he helped Qihoo 360's Vulcan team exploit security vulnerabilities in Microsoft's Internet Explorer browser at the international Pwn2Own hacking competition. He later presented his research at Black Hat Europe, a major cybersecurity conference. In 2019, He became the first Chinese researcher to top Microsoft's annual Most Valuable Security Researchers ranking. He secured first place four more times between 2021 and 2024.
A 2024 study by the Center for Security Studies at ETH Zurich, a leading Swiss university, described him as one of China's prominent civilian "superstar hackers." He has also won two Pwnie Awards, prestigious cybersecurity honors sometimes called the "Oscars of hacking."
Chinese media reported that Chen has joined Z.ai's GLM foundation model team to strengthen its cybersecurity capabilities.
Founded in 2019 by researchers associated with China's prestigious Tsinghua University, Z.ai, formerly known as Zhipu AI, develops AI models that compete with those from DeepSeek and U.S. companies OpenAI and Anthropic. In July 2025, the company introduced an AI model it claimed was cheaper to use than DeepSeek's, with CEO Zhang Peng telling CNBC that it could operate using just eight Nvidia chips.
It became the first major Chinese generative AI startup to go public earlier this year. Its market valuation reached approximately US$32 billion in October, according to Bloomberg, after it raised another $5 billion through share and convertible bond sales in September to fund its expansion and AI research.
It said in August that its GLM-5.3 model had significantly improved its ability to discover software vulnerabilities compared with its predecessor. A subsequent assessment by U.S. AI company Anthropic found that GLM-5.3 demonstrated advanced capabilities in identifying and exploiting cybersecurity weaknesses. Chen said in late September that he was switching to GLM-5.3 to assist his security research after reading Anthropic's findings.
