Chinese Cybersecurity Stocks Surge After Mythos Announcement on April 7

Following the Mythos announcement on April 7, shares of major Chinese cybersecurity firms rose for several consecutive days.

The development stems from the release of Anthropic’s Claude Mythos Preview, an artificial intelligence model designed to autonomously identify and exploit cybersecurity vulnerabilities at a level surpassing conventional enterprise and financial system tools. Although the model is not publicly available and Anthropic’s services remain unavailable in China due to the company’s classification of China as an “adversarial nation,” the announcement has nonetheless triggered significant market reactions within the country’s cybersecurity sector.

According to reports, shares of prominent Chinese cybersecurity firms including Qi An Xin, Sangfor Technologies, and 360 Security Technology experienced sustained increases in value after the April 7 announcement. This rally reflects growing investor expectations that demand for AI-driven cybersecurity and compliance solutions could accelerate in response to the perceived capabilities of models like Mythos.

Austin Zhao, a senior research manager at IDC China, noted that the strong market reaction underscored the sector’s close attention to Mythos. While the model’s advanced capabilities had been anticipated within the industry, Zhao stated that its actual performance still surprised many observers, contributing to heightened optimism about future growth in AI-integrated security offerings.

The Mythos announcement was accompanied by the launch of Project Glasswing, an initiative in which Anthropic provided the model to a consortium of U.S. Companies—including Cisco, JPMorgan Chase, and Nvidia—to detect and remediate vulnerabilities in critical open-source software. The U.S. National Security Agency and other governmental bodies have also begun using the model, according to U.S. Media reports, though no Chinese organizations were included in the initial access group due to geopolitical restrictions.

Despite the absence of direct access to Mythos within China, domestic AI models have shown rapid progress in recent years. Beijing-based consultancy Concordia AI observed that while China’s primarily open-source models still lag behind closed-source U.S. Models in cyber capabilities, their capabilities are improving quickly, suggesting a narrowing gap over time.

The broader implications of Mythos have extended beyond cybersecurity into financial markets globally. In the weeks following the announcement, cybersecurity stocks in the United States experienced significant selloffs, with the iShares Cybersecurity ETF losing 4.5% and individual companies such as CrowdStrike, Palo Alto Networks, and Tenable seeing double-digit percentage declines. Analysts attributed the volatility to fears that advanced AI models like Mythos could disrupt traditional cybersecurity firms by enabling more sophisticated autonomous threat detection and exploitation.

As of April 23, 2026, the Mythos model remains in a controlled testing phase, with access limited to selected organizations under strict supervision. No public release timeline has been announced, and Anthropic has not indicated plans to expand access to entities in greater China, including Hong Kong, amid ongoing restrictions on its services in the region.