Comment by SkyMarshal

> that poses a serious risk to national security, national economic security, or national public health and safety

That seems to be a key component. I imagine many AI companies will start with a default position that none of those are apply to them, and will leave the burden of proof with the govt or other entity.