IBM Partners With OpenAI to Expand Enterprise AI Push
by Deepak Mehra - 5 hours ago - 5 min read
IBM is deepening its push into enterprise artificial intelligence through a new strategic partnership with OpenAI, bringing GPT-5.6, Codex and ChatGPT Work into the systems IBM uses to help large organizations modernize applications, automate workflows and manage cybersecurity.
Announced on August 13, the partnership goes beyond simply giving IBM customers access to OpenAI models. IBM plans to embed OpenAI technology into IBM Consulting Advantage, its AI-powered consulting delivery platform, while creating industry-specific solutions for sectors including financial services, government, telecommunications and retail.
The companies are targeting a problem that has become increasingly important in enterprise AI: businesses have access to powerful models, but integrating those models into decades-old systems and complicated internal workflows remains difficult.
OpenAI's models are moving into IBM's consulting stack
Under the agreement, GPT-5.6, Codex and ChatGPT Work will become part of IBM Consulting Advantage. IBM says the combination could be used across areas such as finance, procurement, customer operations and human resources.
IBM is also creating a dedicated OpenAI Practice, with thousands of IBM consultants and engineers expected to obtain advanced certifications through the OpenAI Partner Network. Specialized forward-deployed teams will work directly with customers, including companies operating in highly regulated environments.
The partnership broadly concentrates on three areas:
| Area | Planned use |
|---|---|
| Enterprise workflows | Use GPT-5.6 and other OpenAI tools to redesign and automate business processes |
| Software modernization | Combine Codex with IBM technology and consulting expertise to update legacy applications |
| Cybersecurity | Combine OpenAI frontier models with IBM Autonomous Security and its AI agents |
IBM Managing Partner Michael Healy said the focus will begin with the business use case rather than simply deploying a model. IBM's Process Studio, for example, can examine existing procedures and identify inefficient steps before AI agents are introduced into those workflows.
Codex could become an important part of the deal
Application modernization may prove particularly significant.
Many of IBM's largest customers operate complex legacy technology estates. The companies plan to combine Codex and ChatGPT Work with IBM's application-development and industry expertise to help businesses update those systems and speed up software development.
That fits a broader change OpenAI is already seeing among corporate customers. As of June 2026, Codex accounted for 64% of the combined output tokens generated by Codex and ChatGPT among OpenAI enterprise customers. OpenAI also reported that weekly active enterprise Codex usage since February had grown 108x in legal, 41x in sales, 41x in recruiting and 26x in marketing, compared with 5x in engineering.
Those numbers suggest enterprise AI adoption is increasingly moving from employees asking chatbots questions toward agents completing longer, multi-step tasks.
Cybersecurity remains another major piece
IBM and OpenAI were already working together before this week's broader partnership.
IBM joined OpenAI's Daybreak Cyber Partner Program in June and introduced an application-security service that uses OpenAI's cyber capabilities to identify and validate software vulnerabilities. IBM said the service can analyze code, prioritize potentially exploitable weaknesses and operate with controlled, read-only access to customers' repositories.
The expanded partnership will now connect OpenAI's frontier AI capabilities with IBM Autonomous Security, IBM's multi-agent system for coordinating security intelligence, decisions and responses. IBM also reported that AI-driven cyberattacks increased 56% in its 2026 security research, adding urgency to the defensive-AI side of the agreement.
IBM has a financial reason to make enterprise AI work
The partnership also arrives as IBM looks for stronger growth from its AI and consulting businesses.
IBM reported $17.2 billion in second-quarter 2026 revenue, up just 1% from a year earlier. Software revenue rose 5% to $7.8 billion, while Consulting generated roughly $5.3 billion and remained essentially flat year over year. IBM nevertheless expects full-year revenue to grow 4% to 5% at constant currency.
IBM has already demonstrated significant commercial demand around AI. Its generative AI book of business had surpassed $12.5 billion by the end of 2025, according to the company.
OpenAI, meanwhile, is pushing aggressively into the same corporate market. ChatGPT Business currently has more than 5 million business users worldwide, while OpenAI said late last year that it had surpassed 7 million ChatGPT workplace seats and that Enterprise seats had grown roughly ninefold year over year.
IBM is not betting on only one AI ecosystem
One interesting detail is that the OpenAI deal does not mean IBM is abandoning its open-model strategy.
Just two days before announcing the expanded OpenAI partnership, IBM signed a $240 million multi-year agreement with Together AI to build an AI inference cluster on IBM Cloud. The initial deployment is expected to use roughly 2,000 Nvidia Blackwell 300 chips and support open models including DeepSeek, MiniMax and Kimi.
That gives IBM a potentially useful position in the enterprise AI race: it can offer customers OpenAI's proprietary frontier technology while continuing to support open models, watsonx and hybrid-cloud deployments.
The OpenAI partnership therefore looks less like IBM choosing a single AI winner and more like an attempt to become the enterprise integration layer connecting whichever models businesses want to use with the systems, data, security controls and workflows they already have.
For IBM, that integration opportunity could ultimately be more valuable than simply providing access to another large language model.