PHP Exposes Two Critical Remote Code Execution Vulnerabilities, Government Websites in Multiple Countries Urgently Taken Offline

Main text

PHP has been hit by an extremely critical vulnerability, government websites in multiple countries have been urgently taken offline, and researchers warn that around the public disclosure of the vulnerability, hacker attack attempts surged from thousands to millions, and global cybersecurity alarms have suddenly sounded.

PHP Extremely Critical Vulnerability Affects the Globe; MIIT NVDB Issues Urgent Risk Alert

On October 21, PHP officially disclosed and fixed two extremely critical remote code execution vulnerabilities (CVE-2026-109618, CVE-2026-109619). Because PHP is an open-source server-side scripting language widely used in website development, and about 70% of websites worldwide use PHP, the vulnerabilities have an extremely broad impact.

After the vulnerabilities were made public, government agency websites in Finland, the Netherlands, Belgium, France, Germany, the United Kingdom, Japan and other countries were urgently taken offline or adopted protective measures. Some Wikipedia systems, some BBC program pages, and the City of Edinburgh Council website were temporarily inaccessible, and the U.S. National Security Agency even advised organizations unable to patch in time to take their systems offline.

It is worth noting that these vulnerabilities were discovered by AI — cybersecurity companies Wiz and Calif used OpenAI's frontier models to discover these vulnerabilities that had existed for more than 10 years, and Wiz called them the "most severe vulnerabilities" discovered by AI.

Domestically, on October 22, the Ministry of Industry and Information Technology's Cybersecurity Threat and Vulnerability Information Sharing Platform (NVDB) issued a risk alert stating that monitoring had found two extremely critical remote code execution vulnerabilities in PHP, which are severely harmful, and advised relevant units and users to immediately carry out comprehensive inspections, promptly upgrade to the latest secure version, or adopt temporary protective measures such as temporary offline shutdown, network isolation, and restricting network access. Cloud service providers such as Tencent Cloud and Alibaba Cloud have also issued vulnerability risk notices, reminding users to check their business systems and install security updates in a timely manner.

Institutions: Cybersecurity Is Upgrading from an IT Cost Item to a Rigid Demand in the AI Era

Cybersecurity refers to the industry that protects network systems, data and applications from attack, destruction and leakage through technical means and management measures.

According to Gartner, the global AI security market is expected to grow 69% in 2027 compared with 2026, reaching US$4.783 billion, and further rise to about US$7.7 billion in 2028.

Open Source Securities recently noted in a research report that as AI applications such as large models and agents accelerate their deployment, attackers have begun using AI to complete target screening, path planning and attack execution, and the rules of the cybersecurity offense-defense game have evolved from person-to-person contests to a stage of "AI versus AI."

In the view of Guotai Haitong Securities, AI security is changing from an "optional configuration" to a "mandatory requirement." The Action Plan for Promoting the High-Quality Development of Internet Information Enterprises (2026–2030) proposes accelerating legislation in emerging fields such as artificial intelligence, improving algorithm governance, and implementing the Regulations on the Administration of Network Data Security, which means AI security is an institutional demand rather than a cyclical demand.

Earnings of Multiple Concept Stocks Expected to Multiply

According to the East Money concept sector, there are currently more than 100 stocks in the A-share market involved in the cybersecurity concept, with a combined total market value of about 4.64 trillion yuan. ZTE, Sugon, Ruijie Networks, Inspur Information and others all have market values exceeding 70 billion yuan.

Since October, the sector has trended weakly. Jilin Zheng Yuan rose more than 35% against the trend to rank first, Huali Chuangtong and Taiji Co. both rose more than 20%, and Nanwei Software, Feitian Chengxin and ST Xintong all rose more than 10%. On the decline side, Ruijie Networks, Yongxin Zhichneg, Tongniu Information and two other stocks fell more than 20%.

From a capital perspective, East Money Choice data shows that since October, a total of 11 cybersecurity concept stocks have received contrarian bets of more than 10 million yuan from margin traders. Among them, Sangfor received 45 million yuan in additional leveraged funds, Huali Chuangtong and Qilin Xinan attracted 40 million yuan and 38 million yuan respectively, and NSFOCUS, Qi An Xin, State Grid Information & Telecommunication and two other stocks each received more than 20 million yuan in net margin purchases.

In terms of future growth potential, according to consensus forecasts from three or more institutions, a total of 10 cybersecurity concept stocks are expected to see their earnings double this year: institutions expect China Great Wall's net profit to surge 11-fold year on year, Yongding Co. and Zhiwei Intelligent's net profits to grow more than 3-fold, and Aofei Data, Pantum Technology, Jinzheng Co., Ruijie Networks and other stocks are also expected to see earnings multiply.

Sangfor is one of the earlier domestic cybersecurity vendors to release a self-developed security large model, "Security GPT." The company builds a full-lifecycle protection system from pre-launch assessment to runtime auditing around AI application protection and agent identity and behavior security. In August, the company's SangforAI rose to first place in the global ranking of the international AI security benchmark CyberGym.

NSFOCUS focuses on AI security, data security and real-world offense and defense, using "Fengyun Wei" to empower security operations and "Qingfeng Wei" to cover the security of large models and agents themselves, forming a product loop from "AI empowering security" to "AI's own security." In the first half of this year, the company's AI security-related order value reached 117 million yuan, of which AI security contract value in the telecom operator industry grew more than 300% year on year.

East Money Illustrated · A Little More Substance

(Source: East Money Research Center) Middle of the main text Embedded advertisement