Artificial Intelligence Driven Approaches to Smart Home Security and Efficiency: A Systematic Review

Abstract

Artificial intelligence (AI) has emerged as a key technology for improving smart home security and energy efficiency. This study aims to systematically review AI-driven approaches in smart home environments. The review follows the PRISMA 2020 framework and uses the Scopus and Web of Science databases, resulting in 2,429 records, of which 60 studies published between 2021 and 2025 meet the inclusion criteria. The findings indicate that AI significantly enhances smart home performance through automation, energy optimization, anomaly detection, intrusion prevention, and predictive decision-making. Machine learning and deep learning are the most frequently adopted techniques, while reinforcement learning and hybrid AI models demonstrate strong capabilities for adaptive control and energy management. The review also identifies challenges related to data quality, model interpretability, privacy, and real-world implementation. Overall, AI plays a vital role in developing secure, intelligent, and energy-efficient smart home systems that contribute to sustainable urban living and smart city development.