Cloudflare Integrates OpenAI's GPT-5.6-Cyber to Automate Network Edge Vulnerability Patching

Cloudflare's Vulnerability Discovery and Remediation service uses OpenAI's GPT-5.6-Cyber via the Daybreak Defense Network to map live request paths to code weaknesses. The system automatically drafts custom firewall rules and code patches for engineers, but requires explicit human approval before deploying the fixes. The National Vulnerability Database logged 60,475 vulnerabilities by September 2026, already surpassing the 48,185 total recorded in all of 2025. OpenAI's GPT-5.6-Cyber model completed 95% of advanced cybersecurity requests on internal benchmarks, vastly outperforming the general GPT-5.6 Sol model's 1.5%. Palo Alto Networks and Proofpoint also integrated Daybreak models into their threat investigation and consulting workflows this month. "A network that reads internet traffic in real time, paired with GPT-5.6-Cyber, lets defenders stop attacks before they land." — Matthew Prince