With the limited availability of comprehensive reviews in existing literature, this systematic review, together with topic modeling, aims to provide a thorough way to give both quantitative and qualitative information on the constantly developing topic of sentiment analysis in financial markets using machine learning (ML) and artificial intelligence (AI). For achieving this objective, this study identifies 723 papers according to “Preferred Reporting Items for Systematic Reviews and Meta-Analyses” (PRISMA) protocol and selects 238 publications from Scopus database for analysis. Furthermore, the Latent Dirichlet Allocation (LDA) analysis is employed to identify the most prominent themes and topics. Subsequently, a thorough examination of the content of 238 papers within the identified clusters is conducted. The four primary topics or themes identified are: sentiment analysis and machine learning in financial markets, advanced techniques in financial sentiment analysis, machine learning innovations in financial market predictions, and role of big data and social media in financial markets’ prediction. We present a concise overview of these topics together with the prospective avenues for further investigation. As AI and ML disrupt the financial industry, continuing research will be needed to monitor trends, evaluate new breakthroughs, and enable stakeholders make informed decisions.