OpenAI is expanding its consulting-like service to refine its models and build bespoke apps for large enterprise and government clients spending at least $10M
As OpenAI pitches its artificial intelligence to big companies and government agencies, it's taking a page from the playbook … X: @amir X: Amir Efrati / @amir : New: OpenAI now does Forward Deployed Engineers too. Ultimate goal: getting a large customers to pay $1B each for customized AI models Former head of Preparedness Aleksander Madry is ~on it~ [image]
Context & Ripple Effects
OpenAI had already been building an enterprise revenue base, with ChatGPT Enterprise reported to have more than 300 paying customers in its earlier enterprise push. This service takes that commercial effort beyond standardized model access toward hands-on implementation for a narrow set of high-spending accounts.
The government focus makes deployment capability—not just model quality—a central part of OpenAI's enterprise offer. It is an early example of a frontier-model provider adopting the operating model associated with OpenAI's later deployment-company expansion.
First-order effects
- Large enterprises and government agencies committing at least $10M can receive direct help refining OpenAI models and building bespoke applications, rather than relying solely on general-purpose products.
- OpenAI must allocate specialized technical staff, including the reported involvement of Aleksander Madry, to a small number of strategic customer deployments.
Second-order effects
- The offer raises the bar for rival model providers seeking comparable accounts: access to capable models alone may be insufficient where customers require customization and on-site deployment support.
- It puts OpenAI in closer competition and partnership with systems integrators and consultancies, which already sell implementation expertise to large organizations.
Third-order effects
- If repeated across major model vendors, enterprise AI competition could shift from broadly available APIs toward a tiered market in which the largest buyers receive tailored models, embedded engineering, and deeper vendor relationships.
- Government procurement may increasingly favor providers that can pair model capability with operational deployment and customization, reinforcing the institutional role of frontier AI labs.
The trend: Frontier AI companies are evolving from software suppliers into deployment partners for a small number of strategically valuable enterprise and public-sector customers.