The first 128 UTF-8 characters precisely match the first 128 ASCII characters (numbered 0-127), meaning that existing ASCII text is already valid UTF-8. All other characters use two to four bytes. Each byte has some bits reserved for encoding purposes. Since non-ASCII characters require more than one byte for storage, they run the risk of being corrupted if the bytes are separated and not recombined.