OpenAI is on a quest, chasing down Anthropic’s Claude toolkit to be seen as the enterprise frontier model. Here’s the push-pull-partner strategy we see unfolding at OpenAI:

  • Push. CEO Sam Altman pushed a vision for an AI-powered world out into the global dialog. Investors, Boards, and The Media loved it. This strategy is alive and well as OpenAI is cited repeatedly by global CEOs as their lead AI supplier. At the same time, on the back of astounding (and well deserved) consumer adoption of ChatGPT, OpenAI started pushing customers to expand how they use the tool, for learning, shopping, content generation, and work. That leads to the pull strategy.
  • Pull. If you use ChatGPT at home, you likely also use it at work Any time a consumer is empowered by technology, they bring it to work. Businesses respond by putting the new technology on the watch list for an enterprise agreement. OpenAI benefited tremendously at first. But it fell behind Anthropic’s Claude Code and Cowork, who did it better in 2025 and 2026. OpenAI and has been digging out of that hole ever since. Which leads us to . . .
  • Partner. When we evaluated the 10 best AI consulting service providers in early 2026, only four of them – Accenture, BCG, Capgemini, and McKinsey – could brag that they were OpenAI’s frontier launch partners. This vaguely defined, but cool sounding arrangement is still a bit of a mystery to us. Despite OpenAI’s quest to build its own deployment force, it needs service provider partnership to scale the technology. The model builder has hired partner managers from Salesforce and Google and announced many partner programs and partnerships since.

The IBM/OpenAI announcement marks a maturing of this strategy. We write this while the ink is still wet on a partnership announcement between IBM and OpenAI. This new partnership will embed OpenAI’s tools into IBM Consulting Advantage (its AI-powered delivery platform) and establish a business group with forward-deployed teams to sell and implement OpenAI-centric solutions. We see this as a signal to the enterprise that Big Blue is now squarely behind OpenAI. That carries weight. This partnership:

  • Checks all the boxes on a tech-service partnership. IBM shifted to a partner strategy a decade ago and now presents its consulting services squarely through the lens of cloud, software, and now AI partnerships. IBM tells us that at least 40% of IBM Consulting revenue is associated with a partner. IBM is setting up a practice and business group with OpenAI-certified engineers and what IBM is calling forward-deployed units, small teams of data science, engineering, and domain expertise. IBM will commit to co-marketing and co-selling OpenAI’s offerings, train its consultants and sales teams, and commit to a sales number.
  • Slots into a partner strategy that ranks among the most aggressive. Why? Because IBM is a product company first and service provider second. Meaning the partnership with OpenAI is also with the product side of the house, with Red Hat, WatsonX Orchestrate, and the ability to secure the solution and incorporate the OpenAI models into an enterprise workflow. This is something that only IBM can do today (though Accenture and HCL are hot on the trail.)
  • Includes a credible enterprise-scale testing ground for OpenAI’s Codex. Anthropic’s Claude Code and Cowork are taking the enterprise by storm. OpenAI is playing catch up. In this partnership, IBM has committed to incorporate OpenAI’s tools into IBM Consulting Advantage, the most advanced AI-powered delivery platform we saw in the evaluation. Because IBM is also a software company, they bring a more sophisticated delivery platform and 130,000 consultants using it. By committing to OpenAI’s technology in this platform, IBM becomes client zero with an enterprise-scale testing ground for software development but also many other enterprise workloads.
  • Targets regulated industries. IBM has committed to securing OpenAI’s solutions with deployment architectures that regulated firms can deploy with more confidence. Based on feedback we get from technology partners like Adobe, Salesforce, and SAP about the IBM partnership, IBM does regulated industries better than other providers. This alone makes the partnership deal unique.

What should technology executives watch for?

  • Take a prove-it-to-me attitude. If this partnership is working, then when you ask IBM Consulting for an improvement in OpenAI’s product, it should show up in OpenAI’s roadmap ASAP. If not, then question why the partnership matters.
  • Don’t make OpenAI your only model. Your enterprise architects are rapidly learning how to optimize AI workloads by parsing the job down into sub-agents that call on different models. IBM can help you with a model routing architecture and platform if you choose. OpenAI’s models are only one ingredient in that mix.
  • Be wary of any offer to write OpenAI technology on IBM paper. In other words, expect that you will need a relationship with IBM and with OpenAI rather than a bundled RFP and price. It’s too soon for “one throat to choke” with AI.

Let Us Know How We Can Help

There is so much going on at the intersection of software and models and services to talk about. Forrester customers can schedule a guidance session to investigate what this new partnership means for you.

Share