Currently only single byte encoded character sets are supported within the request body of HTTP tasks.
Special characters, for example german characters, which include an amulet, can be represented using ISO-8859-1 in the content type request header but it is not possible to specify UTF-8 in the same way and use those characters.
The work around is to write a script in say Python, that performs the HTTP request and which supports UTF-8, but it would be ideal to have UTF-8 support in the dedicated HTTP task.