Recent Advancements and Trends in Artificial Intelligence from the Information Security Perspective: A Systematic Review

The research examines 58 peer-reviewed studies from 2020 to 2025 to study the effects of advanced Artificial Intelligence (AI) techniques on information security (InfoSec). The research divides AI-based security methods into five main security domains which include anomaly detection and intrusion response and privacy-preserving mechanisms and real-time threat intelligence and adversarial resilience. Deep learning and federated learning drive the advancement of adaptive security frameworks and reinforcement learning enables automated policy enforcement and natural language processing (NLP) enhances phishing detection and policy interpretation. The field demonstrates significant advancement but faces essential obstacles because explainable AI adoption remains restricted and benchmarking standards are inconsistent, and ethical and fairness principles receive inadequate attention. The review establishes a fresh classification framework for AI applications in InfoSec and outlines strategic research paths to boost security-critical environment interpretability and scalability and responsible deployment.