CVE-Genie raises the stakes in the vulnerability management race

CVE-Genie is a multi-agent AI framework that automates gathering CVE data, rebuilding vulnerable environments, generating proof-of-concept exploits, and verifying exploitability; in tests it reproduced 428 of 841 CVEs (51%) across many languages and projects. The paper highlights strengths (effective for web-style and interpreted-language bugs, low cost per CVE) and weaknesses (struggles with memory-safety and complex builds), and warns that open-sourcing such tooling could significantly accelerate attacker weaponization and shrink time-to-exploit, requiring faster, intelligence-driven defensive programs.