OpenAI’s latest coding move is aimed not just at helping developers write software, but at the infrastructure that decides what happens to that code once it exists.

The company will acquire Pierre Computer, a developer-tools startup whose technology already powers parts of OpenAI’s Codex platform. OpenAI plans to integrate Pierre’s Code Storage product more deeply into Codex; financial terms were not disclosed, and Pierre’s open-source projects are set to remain open-source.¹

The timing matters. GitHub has long served as software development’s central repository, but it has faced reliability problems and outages as AI coding tools have accelerated. Critics’ broader complaint is that the service was built for human developers, while an increasing share of code is now generated by AI agents.¹

OpenAI has been laying the groundwork. Earlier in the week, it unveiled Pando, a system intended to manage the surge of agent-generated code; the company said it cut the time required to push code changes by nearly 80% and was exploring customer access.¹ Pierre now gives Codex a repository-focused layer that OpenAI says will help “connect the code a company ships with the decisions its people make.”²

That puts OpenAI into a widening contest around GitHub, owned by Microsoft — which also owns nearly a quarter of OpenAI. Other challengers are emerging, including Cursor’s Origin and a product from former GitHub chief executive Thomas Dohmke.¹

For OpenAI president Greg Brockman, the Pierre deal is framed as an extension of an existing partnership: “pierre’s open-source developer tools already power parts of codex today,” he wrote, adding that the companies would build “the future of collaboratively building software with ai.”Greg Brockman (gdb)³

The acquisition sharpens OpenAI’s wager: in the AI era, controlling the place where code is stored, reviewed and shipped may matter as much as generating the code itself.¹