AI-Enabled Personalization and Adaptive User Interface Design in Mobile Banking Applications: Opportunities for Saudi Financial Technology (Iconic Research and Engineering Journals)
Summary
- The paper provides an integrative review synthesizing evidence from 30 peer-reviewed papers published between 2020 and 2025 on mobile banking AI personalization.p.1
- The analysis distinguishes personalization as prioritizing information from adaptation as altering interface structure, presentation, timing, or modality.p.3
- The study organizes collected evidence across 5 core coding domains including utility, trust, security, explainability, and institutional accountability.p.3
- Adaptive user interface design is categorized into 4 operational levels: content adaptation, ranking adaptation, presentation adaptation, and interaction adaptation.p.5
- An integrated design framework is introduced structured around a 5 layer architecture spanning permissioned data to accountable evaluation.p.8
Key insights
- Predictive personalisation and interface adaptation are conceptually separate interventions that must be evaluated independently to avoid overestimating AI influence.p.3
- Low-consequence shortcut adjustments reduce search friction without changing decision paths, whereas high-consequence suggestions introduce complex suitability questions.p.4
- Core transaction steps must maintain navigational continuity to prevent disorienting users who rely on learned safety habits.p.5
- Trust must be calibrated so consumers accept AI prompts when accurate but feel empowered to challenge them when uncertain.p.6
- Explanations should be provided at the point of decision to enable active contestability rather than serving as post-hoc justifications.p.6
- Saudi open banking creates opportunities for aggregated budgeting views if granular consent, data boundaries, and revocation rules are strictly enforced.p.7
- Accessibility features and language modes should be explicitly set by users rather than dynamically inferred by predictive algorithms.p.7
Topics and entities
Summary and figures generated from the document and checked automatically: every number shown appears on the cited page. The full document is available from the publisher. Publishers can request removal.