This isn't just access control—it's OpenAI admitting they've crossed into genuinely dangerous AI territory while trying to arm defenders first.

This is how the AI cold war begins: every country wanting 'sovereign AI' they control, but only a few nations capable of building frontier models.

This isn't another coding assistant. It's the first AI that can work like a human engineer: long-horizon tasks, real-time collaboration, and debugging its own training.

Forget ChatGPT demos. This is AI's real killer app: making physical R&D faster and cheaper through closed-loop automation that delivers measurable ROI.