Microsoft introduced MAI-Voice-1 for fast audio generation on September 2, along with a preview of MAI-1. These tools enhance AI capabilities in voice and audio processing.