Is a "character" in a "string" represented with 2 bytes (16-bits) or simply 1 byte (8-bits)? A 2-byte character type can be used to represent more symbols than a 1-byte character. To represent all characters or symbols of some languages, characters should probably be 16-bits (2 bytes) in length or size.
Based on function str_for_asc found in the manual, I assume that characters in Acknex strings are represented with only 8-bits (1 byte). Also, the functions in the SDK appear to use the standard char data type. May best guess is: Currently, the Acknex Engine has no specific support for 'Unicode'.