Cloudflare launches AI vulnerability service with OpenAI

Cloudflare, Inc. (NYSE: NET) announced a new security service called Vulnerability Discovery and Remediation, now available in early access through its Cloudflare Managed Defense offering. The service uses OpenAI Daybreak models, including GPT-5.6 Cyber, to identify, block, and generate fixes for software vulnerabilities.

The service integrates AI code analysis across Cloudflare's platform, spanning Web Assets, WAF, and Workers Observability. It correlates live internet traffic data with code vulnerability scans to surface high-priority security issues. Custom WAF rules can be deployed at the network edge to block attack vectors before code-level patches are written. OpenAI models also generate code patches for developer review. According to the announcement, no code fix or edge rule takes effect without explicit human approval.

The service is currently available by invitation to select Cloudflare Enterprise customers.

Cloudflare cited data from the National Vulnerability Database showing that 60,475 vulnerabilities had been logged by September 2026, surpassing the 48,185 total recorded across all of 2025.

"We built Cloudflare's global network to analyze what's happening across the Internet in real time," said Matthew Prince, co-founder and CEO of Cloudflare. "Pair that with the power of OpenAI GPT-5.6 Cyber, and you're not just reacting to attacks anymore, you're stopping them before they land."

McCall McIntyre, Head of Global Cyber Partnerships at OpenAI, said the collaboration aims to "give defenders the advantage of frontier AI, safely."

You May Also Be Interested In

• Zscaler to cut 3% staff
• Dutch Bros Inc. (BROS) files mixed shelf
• Korn Ferry (KFY) files for 3.12M share offering by selling stockholders