BEYOND BANDWIDTH: INFRASTRUCTURE CONSTRAINTS, COST BARRIERS, AND AI-ASSISTED SECURE DIGITAL PRESERVATION IN RESOURCE-LIMITED INSTITUTIONS

Abstract

Digital preservation research has traditionally addressed infrastructure and cost as adoption barriers to be overcome on the way to a stable end state. This study argues, using qualitative case-study evidence from a Ghanaian school information system, that infrastructure limitations are not merely a one-time adoption hurdle but an ongoing operational and security condition that resource-constrained institutions manage continuously. Stakeholders across roles described system slowdowns tied to unreliable internet connectivity, the cost of acquiring and maintaining compatible devices, and the need for regular system maintenance and backups, all recurring themes rather than one-off implementation issues. The findings are situated within literature on internet connectivity, digitisation cost, and long-term data accessibility, and it is argued that infrastructure constraints have a direct security dimension: institutions that cannot reliably back up, patch, or monitor their systems are more exposed to data loss and unauthorised access regardless of how well-designed their access controls are on paper. The study closes with a discussion of artificial-intelligence-assisted, low-bandwidth approaches, including lightweight anomaly detection, automated backup scheduling, predictive maintenance, and cloud workload protection adapted for intermittent-connectivity settings, as a promising but unproven direction for helping infrastructure-constrained institutions sustain secure digital preservation without proportionally large capital investment.