IBM and OpenAI have announced a partnership aimed at helping enterprises deploy Artificial Intelligence (AI) across core business operations.
The two companies said the work will cover business operations, application modernisation, software development, security, and AI risk management. The initial focus includes industries such as financial services, government, telecommunications and retail.
According to the agreement, IBM will integrate OpenAI models and products including GPT-5.6, Codex and ChatGPT Work into IBM Consulting Advantage, the platform IBM uses to support consulting and AI delivery work. IBM is also establishing a dedicated OpenAI-focused organisation within its consulting business, with thousands of consultants and engineers expected to receive specialised training and certification.
IBM will also join OpenAI’s Elite Partner tier. The partnership builds on an existing cybersecurity agreement through OpenAI’s Daybreak Cyber Partner Program, announced in June.
There are three broad areas to the agreement:
Enterprise workflows: IBM and OpenAI will work on bringing AI into functions such as finance, procurement, customer service and human resources.
Application modernisation and software development: Codex and other OpenAI capabilities will be combined with IBM’s experience in modernising legacy systems and supervising complex enterprise environments.
Cybersecurity and AI risk: The companies plan to extend their existing cybersecurity collaboration, combining OpenAI’s AI capabilities with IBM security services and tools.
IBM’s Integration, Consulting and Enterprise Controls
IBM’s contribution is primarily the layer around the models: consulting expertise, enterprise systems integration, industry knowledge, governance and security.
Large companies rarely operate on a clean technology stack. Their environments typically combine legacy applications, cloud services, private infrastructure, databases and industry-specific software. An AI model can generate code, summarise information or reason over data, but connecting those capabilities safely to production systems requires additional engineering and operational controls.
IBM has been building its enterprise AI strategy around this deployment layer. Its IBM Consulting Advantage platform aims to help consultants use AI agents and workflows across consulting engagements, while its wider enterprise strategy emphasises orchestration, governance and hybrid-cloud management.
OpenAI’s Models and AI Development Tools
OpenAI brings the model and AI-product side of the partnership. The integration includes GPT-5.6, Codex and ChatGPT Work. Codex is particularly relevant to software development and application modernisation, while ChatGPT Work is positioned for enterprise workplace use.
OpenAI has increasingly built partnerships around this same deployment problem. In June, it launched the OpenAI Partner Network, saying enterprises need more than access to models; they also need help identifying use cases, redesigning workflows, integrating existing systems and managing adoption. OpenAI said it was investing $150 million in the partner ecosystem and aimed to train 300,000 certified consultants by the end of 2026.
[ALSO READ: Sophos Brings OpenAI Frontier Models to MSPs Through Sophos Fusion ]
The Bigger Market Context
The IBM-OpenAI agreement comes as AI companies increasingly work with large consulting and technology-service firms rather than relying only on direct sales.
OpenAI has already announced enterprise partnerships with organisations including BCG, McKinsey, Accenture and Capgemini, with those firms helping customers integrate AI, redesign workflows and deploy systems at scale.
IBM and Anthropic, for example, announced a 2025 partnership focused on enterprise software development, application modernisation, security and governance.
The pattern suggests that IBM is presenting itself as an enterprise integration layer across multiple AI ecosystems, rather than tying its consulting business to a single model provider.
[ALSO READ: IBM and Anthropic Forge Strategic Partnership to Redefine Enterprise AI Development with Security and Governance ]
The most important part of this agreement is the division of responsibility.
OpenAI has the frontier models and AI development tools. IBM brings the enterprise environment in which those technologies have to operate: existing software, security controls, industry processes, data systems and implementation teams.
That division also explains why the companies are not necessarily direct competitors in this arrangement.
The question will be whether IBM can turn access to OpenAI technology into repeatable enterprise deployments, rather than one-off projects.
For OpenAI, the question is whether partnerships with large integrators can shorten the distance between model capability and sustained production use.
