The fact
AI developers can voluntarily submit models for government security testing, but mandatory approval is explicitly ruled out
The true voluntariness of cooperation remains unclear given increasing government pressure on AI companies
Click the link to read an article on the topic: