This is by design, and is NOT a bug. mIRC sends data to and receives data from a server as binary data(since that's all computers actually understand). Before mIRC 7.x it used the local machine's active CODE PAGE to decide how to convert the received data into text or how to convert the text to binary in order to send it to the server. With mIRC 7.x code page support was intentionally removed in favor of utf8*. Due to this any character who's ANSI value would be above 127 is re-encoded by a 2 (or more) byte representation.
*: The move makes sense as most operating systems are trying to move away from code pages in favor of utf/unicode anyways.
Further Reading:
Utf8/Unicode ExplainedWhy code pages were removed from mIRC
