Today at Cyber at the Frontier, OpenAI announced an integration with Endor Labs as part of its Daybreak Defense Factory, which helps teams harden and defend the critical applications we all rely on.
The integration targets a problem that has moved from finding vulnerabilities to fixing them. According to research by J.P. Morgan, new vulnerabilities are being disclosed 16.5 times faster than teams can patch them. Almost 80% of exploits now happen on or before the day of disclosure, and in roughly 60% of breaches a patch was already available.
Closing the gap requires both the reasoning and automation to address vulnerabilities at scale, and the application context to know which ones matter and how to fix them. The OpenAI Daybreak Defense Factory brings together cyber reasoning models, workflows, and technology partners to tackle this challenge. Endor Labs contributes real-time vulnerability intelligence, reachability analysis, and remediation guidance to help defenders prioritize real risk and accelerate fixes.
Bringing Endor Labs into the Daybreak Defense Factory
Endor Labs now connects to Codex Security via the Daybreak Defense Factory. Each Endor Labs finding arrives in Codex with supporting evidence: full-stack reachability, risk, impact, and remediation guidance. Customer-defined policies help prioritize findings based on the risks that matter most to each organization.
With that evidence, the Defense Factory can take a finding from detection through remediation: Endor Labs prioritizes the risk, Codex implements and tests the fix, and Endor Labs verifies the result against policy before human review.
“With Daybreak Defense Factory, security teams can bring findings from Endor Labs together with context from their other tools to investigate vulnerabilities and prepare fixes. This helps defenders act on the findings they already have, using the tools they already use.”
- Matt Boyle, Head of Product Engineering for Cyber, OpenAI
Faster, more efficient fixes inside the Daybreak Defense Factory
Our teams have worked closely together to help more organizations prioritize and fix open source vulnerabilities.
- Real-time vulnerability intelligence. Endor Labs has scanned more than 2 trillion lines of open source code and continuously screens more than 50,000 packages per month for malicious behavior. This gives Codex up-to-date intelligence on vulnerable functions, emerging threats, and malicious dependencies, including risks that may not yet have a CVE.
- Reachability. Full-stack reachability and exploitability across direct and transitive dependencies, attached to each open source finding in Codex. Agents work the vulnerabilities your application can actually hit and skip the rest. That cuts open source findings by up to 99%.
- Context agents don’t have to rebuild. The code context graph tells an agent how the application uses each dependency, so it starts from evidence instead of re-reading the repository. In Endor Labs benchmarks, agents completed security tasks 19x faster with 92% fewer tokens.
Codex uses that evidence to implement and test a fix. AURI Agents provide specialized security analysis and remediation guidance across first-party code, open source, and containers.
- Source code. The AI SAST agent gives Codex the evidence and remediation guidance for vulnerabilities in first-party code.
- Open source. The SCA Remediation agent gives Codex the reachability evidence and remediation guidance for each open source vulnerability.
- Container images. Endor Labs gives Codex the reachability evidence and guidance to remediate vulnerabilities in container base images.
See it in action
Get started
To get started with Endor Labs in your OpenAI Daybreak Defense Factory, follow the import guide.
What's next?
When you're ready to take the next step in securing your software supply chain, here are 3 ways Endor Labs can help:
