What happened?

IBM announced on Thursday that it has entered a partnership with OpenAI. The deal is aimed at bringing OpenAI's models and tools to more enterprise customers through IBM's global consulting business. Financial terms were not disclosed.

The two companies will jointly market AI offerings and develop industry-specific solutions for sectors including financial services, government, telecommunications and retail. A dedicated OpenAI practice is being established inside IBM Consulting; according to Mike Healy, managing partner at IBM Consulting, tens of thousands of consultants — mostly through the retraining of existing employees — will be trained and certified on OpenAI's technologies over the coming months.

What the training covers

By Healy's account the training falls into four areas: OpenAI's Codex tool, the API, cybersecurity and consultative solution credentials. On top of that, a group of "Forward Deployed Experts" will be built up through OpenAI's Partner Network.

On the product side, GPT-5.6, Codex and ChatGPT Work will be integrated into IBM Consulting Advantage, the AI platform IBM built for its consultants. The aim is to speed up how clients deploy AI across core business operations.

Why does it matter?

The agreement is the latest link in OpenAI's effort to grow its enterprise business through consulting firms and technology partners. The company has previously announced partnerships with IT services firms including Infosys and Tata Consultancy Services.

The trend here is more interesting than any single deal: competition among model developers is shifting from building more capable models towards winning corporate customers and large-scale deployments. Who carries the model into the organisation is becoming as decisive as the model itself.

The picture on IBM's side

For IBM the deal widens the range of its frontier AI partnerships. The company pursues a model-agnostic strategy that combines its own Granite family with third-party offerings, and has increasingly positioned itself as an integrator of multiple models through its watsonx platform and global consulting business.

The timing is telling too: IBM lowered its full-year revenue forecast in July 2026 after weaker-than-expected quarterly results. On its most recent earnings call, chief executive Arvind Krishna maintained that AI remains a long-term growth driver and said AI adoption was complementing, rather than replacing, demand for the company's mainframe business.

Not the first partnership

The two companies worked together in June on the cybersecurity-focused OpenAI Daybreak Cyber Partner Program. The new agreement extends that relationship, integrating OpenAI's models with IBM Autonomous Security, the company's multi-agent cybersecurity service.

In figures

  • Consultants to be trained: tens of thousands, mostly by retraining existing staff
  • Training areas: Codex, the API, cybersecurity, consultative solution credentials
  • Target sectors: financial services, government, telecommunications, retail
  • Products to be integrated: GPT-5.6, Codex, ChatGPT Work
  • Financial terms: not disclosed

There is one more piece of context: less than a year ago IBM announced a similar alliance with Anthropic. IBM is therefore carrying models from two frontier laboratories to its enterprise customers at the same time — that is what its model-agnostic strategy looks like in practice.