OpenAI's acquisition of applied-AI consultancy Tomoro has prompted a useful round of strategy questions across the AI market. The deal is small in dollar terms by frontier-lab standards but consequential in what it says about where OpenAI thinks the long-term margin of enterprise AI is going to settle.
The simple read is that frontier labs are no longer satisfied with selling raw model access into someone else's services business. Implementation is where the bulk of customer spending lands. Large enterprises typically spend several times more on the consultants and integrators who deploy AI than they do on the model itself. Owning a credible piece of that layer changes the economics for any lab that can pull it off.
The consulting layer is the prize
Consulting firms have been among the largest beneficiaries of the AI boom so far. The big four professional services firms, the dedicated technology integrators and a wave of AI-native consultancies have all expanded headcount and pricing in line with corporate demand. For a frontier lab, that revenue is sitting in customer accounts that already buy models, and could in principle be captured directly.
The hard part is talent and culture. Consulting businesses scale on people, partnership models and senior client relationships, not on engineering productivity. Several previous attempts by software vendors to acquire consultancies have struggled with the retention of senior partners and the cultural fit with a product-led parent.
What the deal changes
OpenAI's strategy now reads more like a vertically integrated AI provider than a wholesaler of frontier models. The combination of frontier models, an enterprise unit, an applied-AI consultancy and a new vehicle reportedly named The Development Company gives the lab a much larger surface to engage with very large customers.
The competitive implications are not trivial. Anthropic has pursued a partnership-led path with Microsoft, Goldman Sachs, Blackstone and General Atlantic. Google has invested in its own consulting capacity alongside Vertex AI. Microsoft has its own enormous consulting partner ecosystem on top of an aggressive direct sales motion. Each lab is now testing a different theory of where AI margin will eventually accumulate.
Regional considerations
For Gulf-based buyers, the practical question is which combination of model and integrator they want as their long-term partner. Regional system integrators have grown rapidly, and several have begun bundling frontier-model capabilities into managed services priced in dirhams or riyals. Direct engagement with frontier labs remains an option, but mostly for the largest customers with enough scale to justify a bespoke contract.
The deal is unlikely to be the last of its kind. Expect more frontier labs to fold implementation capacity inside their own walls in the next twelve months.
