Over 100,000 organizations rely on security products through Check Point's AI, workspace, cloud, and network security infrastructure. Protecting that architecture against increasingly sophisticated cyber threats requires translating raw vulnerability intelligence into functional defenses before attackers can exploit the gaps. As Jonathan Zanger, CTO of Check Point, noted, the core operational bottleneck in this process has long been human capacity rather than computational power. When a new vulnerability emerges, security teams face a race against time to investigate the flaw and build reliable indicators of compromise.
Historically, this research workflow took hours to complete because the team's output was strictly constrained by the number of analysts available to parse the data. To break through these capacity limits, Check Point deploys OpenAI Daybreak to automate the core of this workflow. By giving defenders access to specialized cyber models that analyze vulnerabilities, reason through attack paths, and identify security gaps in code and configurations, the company reduces research work from hours to minutes.
Automating Threat Research Workflows
1
The economic rationale for deploying advanced cyber models in enterprise security lies in compressing the latency between threat discovery and signature deployment. Check Point Research investigates emerging vulnerabilities as part of the company's prevention-first approach, translating findings directly into customer protections. Daybreak accelerates that investigative work by automating the parsing of complex code configurations and uncovering how systems could be exposed to attack. Jonathan Zanger stated that the company uses Daybreak to speed up and process these tasks in minutes instead of hours, unlocking significant operational efficiencies.
OpenAI Daybreak can understand vulnerabilities, analyze different attack patterns and scan the code to identify gaps.
This capability alters the unit economics of vulnerability research by shifting the analyst's role from manual investigation to validation and orchestration. Can autonomous security models maintain analytical precision and safely manage critical enterprise infrastructure without introducing new attack vectors or blind spots? Check Point addresses this operational risk by defining strict boundaries for autonomous execution, requiring human approval for consequential actions while the system explains its findings and recommended actions.
Scaling Toward Autonomous Security Orchestration
With vulnerability research now taking minutes to complete, the operational focus shifts directly toward remediation and automated orchestration. Check Point is working with OpenAI to develop its Agentic Network Security Orchestration solution using Daybreak, combining continuous monitoring of customer environments with clearly defined operating boundaries. In this architecture, customers define their security objectives while the autonomous system processes telemetry and presents recommended actions for human approval.
Combining OpenAI’s Daybreak with our existing security workflows unlocks a level of automation that can prevent threats in real time.
As security vendors race to deploy autonomous defense pipelines across industrial environments, the pressure mounts on traditional manual workflows that cannot scale against the sheer volume of modern exploits. With Daybreak reducing vulnerability research from hours to minutes across more than 100,000 protected organizations, the bottleneck in enterprise defense has officially shifted from human analysis speed to the velocity of automated remediation.
What changed
Compressing vulnerability research from hours to minutes moves the main security bottleneck from flaw detection to the speed of rolling out defensive measures. Check Point is deploying this engine for its researchers while strictly constraining the autonomy of the system with mandatory human sign-off for critical actions. As a result, the rapid gain in analytical speed inevitably collides with the operator's ability to review and approve recommendations. Threat analysis now takes minutes, yet the final remediation still waits on a human signature.
Our stories: OpenAI Launches Daybreak Cybersecurity on Amazon Bedrock (August 17).