Soil–Forage–Livestock systems (SFL-systems) integration is fundamental for sustainable land management in arid lands, where conventional crop production is often unfeasible. Aridisols dominate dryland agroecosystems and their edaphic constraints, together with climatic limitations, constitute a major bottleneck for fertility and productivity in key arid regions worldwide. This narrative review provides a taxonomic and edaphic framework to guide sustainable SFL-systems and integrates current approaches and technological applications for forage production in arid environments, focusing on an edaphic-digital scheme that combines organic and inorganic soil amendments with AI-based decision support to improve Aridisols productivity and resilience. Searches of the literature (ScienceDirect, EBSCOhost, Clarivate Web of Science; English, 2021–2025) screened 309 records and selected 169 references; seminal older works were consulted for context. Representative quantitative outcomes from the reviewed literature include SOC increases of ~15–30% after multi-year organic inputs; forage biomass gains of ~10–25% following amendments that correct sodicity; and water-productivity improvements up to ~30% with hydrogels or biochar. AI tools can improve soil diagnostics and amendment selection (diagnostic accuracy improvements of ~15–30% in recent studies) and generate predictive models of amendment–response that facilitate optimization of application rates and water use. The novel contribution of this review is the explicit linkage of SFL-systems and amendment-based soil restoration with AI-driven diagnostics and decision support, providing actionable metrics and research priorities to translate digital diagnostics into measurable forage gains in arid and semi-arid regions. Overall, the evidence suggests that targeted soil restoration, reinforced by AI-based support systems, is a feasible strategy to increase forage availability and ecosystem service provision in drylands.