Additive manufacturing (AM) has progressed from a prototyping-oriented technology to an emerging class of digitally enabled manufacturing systems with growing industrial relevance. However, despite substantial advances in process capabilities, its transition into robust, large-scale production remains uneven and highly constrained. This systematic review critically examines peer-reviewed AM literature published between 2010 and 2025, with a specific focus on process modeling, manufacturing system integration, and industrial deployment. The analysis reveals a clear evolution from early empirical process tuning toward physics-based, data-driven, and hybrid modeling approaches aimed at improving process stability, defect control, and repeatability. While these developments have significantly enhanced scientific understanding, their industrial impact has been limited by high computational cost, lack of model generalizability, fragmented monitoring frameworks, and persistent reliance on post-build inspection for certification. The review further demonstrates that industrial scalability has been driven more by system-level innovations—such as multi-laser platforms, in-situ sensing, and hybrid manufacturing lines—than by isolated process improvements. Nevertheless, these complex systems often introduce new vulnerabilities related to thermal management, control integration, and economic viability. Overall, the findings indicate that the primary barriers to widespread AM adoption are no longer rooted in process feasibility, but in system-level reliability, certification readiness, and integration within established manufacturing ecosystems. This review concludes that future progress in AM will depend on a shift from process-centric optimization toward system-engineered solutions that align modeling, monitoring, control, and manufacturing economics within an industrial context.