OpenAI engineers use Codex Ultrafast to cut incident investigations from hours to minutes
OpenAI engineers are using the Codex Ultrafast model to investigate incidents in minutes, applying it to faster security analysis, code refactoring, and routine developer workflows. The report highlights practical gains in speed for operational and security tasks.