The reason is they keep making public statements like:

> OpenAI is committed to ensuring that the benefits of AI are broadly accessible.

They can't claim that then simultaneously work to keep their cybersecurity models out of reach for non-US citizens like myself.