It was discovered that the ICU unicode library performed incorrect processing of invalid multibyte sequences, resulting in potential bypass of security mechanisms.