OpenAI to acquire Pierre Computer Company to strengthen Codex

OpenAI says Pierre Computer Company will strengthen Codex's capabilities for building, understanding and reviewing code. Deal terms were not disclosed. The acquisition targets the stages of the developer workflow that come after code generation: comprehension and review, where agents produce more code than humans can read.

The move comes during a busy week for coding agents. xAI's Grok Build coordinates up to eight terminal agents. DeepSeek open-sourced Super-Agent, a prototype that picks up coding tasks from GitHub. Microsoft shipped Agent Framework skills for coding assistants. Anthropic says Claude wrote more than 90% of its own code in September. Code review is the bottleneck all of these products share.

There is friction among developers. OpenAI's new shared usage limit makes ChatGPT and Codex draw from the same quota, which has drawn complaints. Community threads also accuse OpenAI of quietly re-enabling model-training telemetry in Codex chats, which raises intellectual-property concerns for companies. An acquisition that improves Codex review may help win back enterprise trust, but only if data controls are clear.

What to watch: how quickly Pierre's technology ships inside Codex, and whether it targets enterprise code-review workflows already served by GitHub and others.