Cloudflare and OpenAI have launched Vulnerability Discovery and Remediation (VDR), an invitation-only early-access service within Cloudflare Managed Defense, that uses OpenAI Daybreak models to automate vulnerability detection, prioritization, and edge-based patching while keeping changes human-approved.

Access is by invitation for selected enterprise customers, and the service is positioned as part of Cloudflare’s managed defense rather than a mass-market product.

Findings are validated and given a risk rating, with high-priority, production-impact issues escalated for quicker attention.

AI-generated patch suggestions are provided, but every proposed change requires explicit human approval before deployment, ensuring no code or edge rule is applied without consent.

Remediation options include immediate stops via custom firewall rules scoped to specific methods or paths, and draft code patches generated by the models for engineers to review.

The service adds contextual evidence by correlating findings with live production traffic, routes, and existing protections, enabling prioritization based on real exposure and defenses.

The move responds to rising vulnerability counts and faster attacker speed, highlighting the challenge for security teams and the goal to reduce workload with AI-guided guidance.

Temporary edge protections, such as custom WAF rules targeted to attack paths, can be deployed while permanent fixes are developed.

The approach aims to shorten the vulnerability exposure window by integrating detection and mitigation within the same platform that handles web traffic and edge security.

The solution contrasts automated, edge-based defense with manual patch-by-patch responses, aiming to reduce security team burnout and improve defense posture.

CEO statements frame the tool as proactive defense that analyzes Internet-wide activity in real time using GPT-5.6 Cyber, shifting from patch chasing to automated protection.

Cloudflare leverages its global network to provide real-time signals for faster identification of weaknesses and potential zero-day exploits, aiding threat intelligence and rapid response.