IT Home On October 28, the tech media site gigazine published a post today (October 28) stating that developer @0x_Osprey created a web page CVE Tracker. By using visual charts, it tracks the number of vulnerabilities with CVE (Common Vulnerability Identifier) numbers obtained, clearly showing the steep upward curve in vulnerability reports since 2026.
IT Home note: CVE is a widely used vulnerability numbering system globally. It is commonly used by security researchers, software manufacturers, and enterprises to track, fix, and report security flaws. CVE Tracker organizes these public records into trend charts, and users can also view the release dates of mainstream AI models, thereby observing the relationships between them and the peak periods of vulnerability discoveries.
Reports indicate that AI companies often use test versions for vulnerability discovery and reporting before the model is officially released. Therefore, the time when the model goes live may not be completely synchronized with a surge in the number of vulnerabilities.
The media said that AI is becoming an “amplifier” for security research, allowing for faster code scanning, replication of abnormal behaviors, and report preparation, thus bringing under inspection areas of software that are difficult to cover by human effort.
But the “more vulnerabilities being discovered” does not necessarily mean the system “is more dangerous”. More reports may expose real risks, or they may contain low-quality, repetitive, or difficult-to-verify vulnerability reports, thereby increasing the triage pressure on open-source maintainers and security teams.