It would be nice if the readFile basic step could recognize the file's encoding. (Writing a file with writeFile can store a BOM at the beginning to help in recognizing the character set later, but this must be only a non-default option to preserve backward compatibility with earlier groovy code reading the created files. This agrees with a rule of thumb "accept liberally, produce conservatively").
There is no need to reinvent the wheel. Using (and white-listing) groovy.util.CharsetToolkit can make this easy.