Anthropic's Mythos AI model boosts Chinese cybersecurity stocks

On April 7, Anthropic unveiled its Claude Mythos Preview, leading to a surge in share prices for prominent Chinese cybersecurity companies such as Qi An Xin, Sangfor Technologies, and 360 Security Technology. This model autonomously detects and exploits cybersecurity weaknesses more effectively than traditional tools, although it remains restricted to select organizations for controlled testing and is not available to the public. Despite Anthropic's services not being offered in China, the announcement has ignited conversations regarding its potential effects on the cybersecurity and finance sectors in the country. Austin Zhao, a senior research manager at IDC China, remarked that the market's response indicated significant interest in Mythos, whose performance exceeded expectations. Additionally, James Gong, legal director at Bird & Bird, suggested that this technology might increase cybersecurity expenses as companies invest more in personnel and advanced defense systems while also paving the way for new AI-driven security services. In a related development, 360 Security Technology reported creating an AI-based 'vulnerability discovery agent' that uncovered hundreds of previously unidentified flaws, including those in Microsoft Office, as highlighted by the Chinese research organization Natto Thoughts.