Actually I found that many maths departments in the US don't really still have the foreign language requirement, even though they still claim to on their website. In the UK there is currently no such requirement (I'm not sure if there ever was). I don't know about other countries.
It is certainly increasingly the case that most papers are available in English, but still by no means all (in particular, plenty of maths papers are still published in French and Chinese, and the volume of papers produced these days means that people certainly don't have time to provide quality translations of them all!).
There is no need to become proficient in another language before starting graduate school, and I believe that you will probably be given some kind of help while in graduate school, or you can at least find out what the other students are doing. Where I did my PhD (in the UK), they had courses in "Russian for scientists".
So in summary, I don't think it's really anything to worry about. I believe the exam (in the US) usually consists of translating part of a maths paper, and you are allowed to use a dictionary (although probably not a maths-specific one?).
Could someone more familiar with the US system please correct me if I'm wrong about any of this?