IBM teams with OpenAI to boost enterprise cybersecurity tools

IBM moved its AI strategy further into enterprise security on June 22, joining OpenAI’s Daybreak Cyber Partner Program and folding frontier models into tools meant to spot software weaknesses faster. The deal underscored a wider shift in commercial AI: the biggest near-term money is not in consumer chatbots, but in systems that can reduce cyber risk, speed remediation and fit inside tightly controlled corporate environments.

IBM said the new application security service will use OpenAI’s cyber capabilities to identify and validate vulnerabilities with greater speed and precision. The company said the tools are designed to be deployed defensively inside enterprise workflows and clients’ environments, a crucial detail for banks, hospitals, insurers and other regulated industries that need AI without losing control of sensitive data and internal systems. IBM’s cybersecurity chief, Mark Hughes, said the partnership expands access to advanced AI capabilities so clients can “surface the most relevant risks faster and act with confidence.”

The move built on Project Lightwell, which IBM and Red Hat announced on May 28 as a $5 billion commitment backed by more than 20,000 engineers. IBM and Red Hat described Lightwell as a way to create a trusted enterprise clearinghouse for open source software and secure the software supply chain from upstream development through production. That made the OpenAI pact less like a standalone product launch and more like a new layer on top of an existing security platform designed to help organizations review code, validate vulnerabilities and close gaps before attackers can exploit them.

OpenAI framed Daybreak around authorization, human judgment, monitoring, safeguards and collaboration with the security community, signaling that the program is meant for controlled defensive use rather than consumer-facing experimentation. OpenAI also said its GPT-5.5-Cyber model scored 85.6% on CyberGym, compared with 81.8% for GPT-5.5, giving vendors a measurable reason to test frontier systems in security operations. The Daybreak partner ecosystem already includes Sophos, Darktrace, Proofpoint, Cato, NCC Group, TrendAI and Tenable, showing that IBM is entering a crowded but fast-forming market.

Investors noticed. Reuters reported that IBM shares rose 3.6% in after-hours trading after the announcement, an early sign that Wall Street sees security-focused AI as a high-value business line. For IBM, the appeal is clear: it can pair long-standing trust with enterprise customers and a heavy security footprint with OpenAI’s most advanced models, then sell AI as a control layer for defense, not just another assistant.