IBM said Monday that it has joined OpenAI's Daybreak Cyber Partner Program, a move the company says will help bring frontier AI capabilities into enterprise security operations. The partnership is aimed at helping organizations respond more quickly to threats that are increasingly automated and fast-moving.
As part of the announcement, IBM also introduced a new application security service built with OpenAI's cyber capabilities. The service is designed to help companies identify and validate software vulnerabilities faster than traditional code-scanning approaches, with IBM saying the tool can prioritize the areas most likely to contain flaws or exploitable paths.
IBM said the new offering is meant to operate inside a client's own environment, using read-only access to code repositories and bounded execution. The company said that setup is intended to keep the analysis controlled, secured and governed while allowing large-scale assessment of application risk.
The service is being delivered as a managed offering, which means customers can start with targeted reviews of specific applications and then expand to ongoing monitoring as code changes and threats evolve. IBM said that approach is intended to support both one-time evaluations and continuous reassessment over time.
The announcement builds on Project Lightwell, a separate IBM initiative focused on software supply chain security. IBM described that project as combining an enterprise security clearinghouse with a global engineering effort to patch, validate and manage open source code. The company said the initiative will draw on OpenAI's cyber capabilities, along with other frontier AI models, to support code review and remediation. IBM said the broader effort is backed by a $5 billion commitment from IBM and Red Hat.
IBM Consulting Global Managing Partner for Cybersecurity Services Mark Hughes said in a statement that attackers are already using AI to probe and scale threats, and argued that defenders need similar capabilities, but with enterprise-grade security and control. He said the Daybreak program broadens IBM's access to advanced AI tools that can be deployed within customer environments to help surface relevant risks sooner.
OpenAI Chief Information Security Officer Dane Stuckey said security is essential to unlocking the value of advanced AI. He said the partnership with IBM and other companies is intended to accelerate defensive security workflows and help enterprises, governments and other organizations identify risks and strengthen resilience while keeping the trust, controls and compliance their environments require.
IBM said the new application security service is available now, with more integrations planned through the Daybreak Cyber Partner Program. The company framed the move as part of a broader effort to shape how frontier AI is used across enterprise workflows, especially in areas where controlled analysis and secure deployment are essential.
The partnership comes as major technology companies look for ways to apply generative AI and related models to cybersecurity, a field where both attackers and defenders are increasingly using automation. IBM's latest announcement suggests the company sees AI not only as a way to speed up detection, but also as a tool for validating the real-world risk of software weaknesses before they are exploited.