Muscle-invasive bladder cancer (MIBC) results in the spread of malignancy beyond the inner bladder lining and into the bladder muscular wall. MIBC is generally considered high risk, requiring aggressive treatment such as radical cystectomy, chemo and radiation therapy, or immunotherapy. However, current advancements in artificial intelligence (AI) have demonstrated increasing accuracy in diagnosis and treatment planning capabilities. Specifically, AI demonstrates high accuracy in cystoscopic cancer detection and differentiation from muscle-invasive or noninvasive disease using deep learning models. In treatment planning, AI has capabilities of making predictive lymph-node models, precise delineation of tumors, and judgements on neoadjuvant chemotherapy outcomes. Yet current challenges such as dataset heterogeneity and regulatory uncertainties continue to exist, prompting future improvements on developing interpretable AI models with interdisciplinary collaboration and clinical validation.