Could the ideological layer also create a security issue by steering user answers in a certain direction?
j previous speech k next speech