Vestibular Schwannoma (VS) is a benign tumor arising from the Schwann cells of the vestibular division of the eighth cranial nerve. This systematic review aims to evaluate the current landscape of Machine Learning (ML) / Deep Learning (DL) applications for preoperative prediction in VS. A comprehensive literature search was conducted, following the PRISMA 2020 guidelines across four databases: PubMed, Scopus, IEEE Xplore, and the Virtual Health Library (VHL). The included studies covered diverse goals, such as tumor growth forecasting, Koos grade classification, screening from audiometric data, and treatment decision support. MRI data, particularly contrast-enhanced T1-weighted scans, were the most commonly used modality. Although several models demonstrated good performance and innovative methodologies, generalizability remains limited due to retrospective, single-center data and lack of external validation. This review highlights both the potential and the current limitations of ML/DL in VS care, emphasizing the need for more robust, interpretable, and clinically integrated models.