I am 4ra1n, an independent security researcher
Currently unaffiliated with any company or organization
我是 4ra1n,独立安全研究员,目前无工作
My Blog: https://4ra1n.github.io
My CVE and Acknowledgments: SECURITY-en.md
我获得的 CVE 和致谢信息记录:SECURITY-zh.md
I see 2025 as a turning point in AI’s cybersecurity capabilities
Before then, I discovered 54 CVEs through manual research
I discovered a total of 100+ CVEs in well-known products (some are in disclosure process)
The products include httpd/tomcat/jetty/mysql/jdk/kafka/activemq/weblogic etc.
我认为 2025 年是 AI 安全能力发生质变的拐点
在 2025 年以前的古法时代我发现了 54 个 CVE 漏洞
我共计发现了 100+ 个知名产品的 CVE 漏洞(部分在发布流程中)
涉及到的产品有 httpd/tomcat/jetty/mysql/jdk/kafka/activemq/weblogic 等



