Unfortunately, the § is at codepoint U+00A7, which is outside the range that UDL officially supports (it is only guaranteed to work with U+0020 to U+007F). There have been at least two issues (#6014 and #8562) requesting Unicode support in UDL, but UDL doesn’t get much attention, so they remain open.
As a workaround, you can add extra highlighting to a builtin lexer or to your User Defined Language (UDL) using regexes via the script EnhanceAnyLexer.py that @Ekopalypse shares in his github repo. Since the matching is done on the Python side, it has full Python 2.7 support for Unicode strings, so you should be able to define your regex inside that script like u'§\w+§' or maybe r'§\w+§'