Palo Alto's NOVA Discovers 14,090 Vulnerabilities in Open-Source Projects

Palo Alto Networks' NOVA AI engine identified 14,090 previously unreported vulnerabilities across 3,915 open-source projects over a two-month period. The findings were validated through a rigorous pipeline, revealing that only 85 of these matched existing records, most of which were published shortly after NOVA's discovery. The vulnerabilities span various types of software, highlighting potential risks in widely used open-source code. The company announced these findings at the Black Hat conference alongside the release of PAN-OS 12.2 Ceres, which offers network-level virtual patches to mitigate risks before official vendor fixes are available. This proactive approach aims to enhance security for organizations relying on open-source software. The NOVA system's capabilities underscore the importance of automated vulnerability detection in the cybersecurity landscape.

Key Points: • NOVA identified 14,090 vulnerabilities in 3,915 open-source projects. • Only 85 of the findings matched previously documented vulnerabilities. • PAN-OS 12.2 Ceres provides rapid virtual patches for discovered flaws.