Heiderich, Tremblay and others have been working on making WeBWorK work well with characters from all languages (see the pull requests on github.com). Release 2.14 should see good progress, if not perfection, in this direction. We'll make an effort insure backward compatibility or at least an easy upgrade path for problem files that were not saved using utf8 encoding.
I believe the problem arises because the file is being saved with a latin character set and the copyright character is different in the latin character set and in the unicode (or utf8) character set that allows for other fonts such as Cyrillic or Hungarian. I don't have an immediate fix other than making sure the file is saved from an editor which is using the utf8 character encoding.