Qi An Xin Targets AI-Driven Defenses Amid Rising Cyber Threats

As malicious actors increasingly leverage artificial intelligence to automate cyberattacks, Beijing-based cybersecurity leader Qi An Xin is scaling its defense capabilities to protect Chinese digital infrastructure. The rise of AI has significantly lowered the barriers to entry for attackers, allowing for faster, larger-scale intrusions that compress the timeframe available for security teams to respond.
Chen Huaping, vice-president of Qi An Xin, highlighted the gravity of the evolving threat landscape at a security conference in Xiamen last week. Following recent global discoveries of critical flaws in the PHP programming language—which reportedly affected millions of assets in China—Chen warned that failing to patch known vulnerabilities now carries costs multiplied by hundreds or thousands of times compared to pre-AI standards. According to data from the U.S. cybersecurity firm Censys, as of Friday, China had patched 5.8 per cent of discovered PHP vulnerabilities, compared to 33.9 per cent in the U.S.

Founded in 2014, the Shanghai-listed company serves approximately 400,000 major organizations, including over 98 per cent of China’s central government departments and major state-owned enterprises. To address the new challenges, the firm has adopted an “All in AI” strategy centered on its AI security subsidiary, led by CEO Gong Yushan. The subsidiary’s dual focus is on using AI to defend clients and securing the AI systems that clients themselves operate.
Central to this strategy is the AI security operations platform, Aisoc. By utilizing the company’s in-house QAX-GPT model, Aisoc can assess internal security alerts in approximately nine seconds, a process that takes human analysts roughly five minutes. In a case study involving a provincial power-grid client, the platform successfully processed more than 10,000 daily alerts, vastly outperforming human capacity.

Competitive pressure remains high, with industry rivals like Qihoo 360 and Sangfor Technologies also expanding their AI-integrated threat detection and analysis offerings. While Huatai Securities analysts noted in a November report that Qi An Xin’s AI-related orders grew 223 per cent in the first nine months of the year, the company reported 2.27 billion yuan in total revenue for the period, a 20.2 per cent year-on-year decline.
The cybersecurity environment remains volatile. Qi An Xin’s threat intelligence team reported a 51.9 per cent year-on-year increase in vulnerabilities discovered during the first half of the year, with 30.2 per cent of high-risk vulnerabilities exploited on the day they were identified. This includes evidence of AI-generated malware, such as that attributed by the company to an unidentified group known as “UTG-Q-022” that targeted a Chinese shipbuilding entity. Furthermore, company researchers recently identified an instance where a jailbroken DeepSeek open model was used to attempt breaches against multiple organizations, including a European telecommunications provider.

As the industry debates the role of open-source models—which some argue bolster security through shared defense while others contend provide tools for attackers—Gong Yushan warned of the narrowing window for defense. “If defenses cannot evolve as quickly as attacks, openness itself will become a source of risk,” Gong said.