Frontier AI models identify exploitable vulnerabilities in 4 hours versus 60 days for human researchers

According to CSO Online, frontier AI models can autonomously identify exploitable vulnerabilities in networks and systems in approximately 4 hours, compared to 60 days required by human security researchers, compressing the window for remediation and raising the stakes for enterprise security operations.