ENHANCING SQL INJECTION ATTACK DETECTION USING MACHINE LEARNING : A REVIEW

SQL Injection (SQLi) remains a critical cybersecurity threat, enabling attackers to exploit database vulnerabilities and compromise sensitive data. Traditional security mechanisms, such as Web Application Firewalls (WAFs) and signature-based Intrusion Detection Systems (IDS), have struggled to counter evolving SQLi techniques. This review explores the integration of artificial intelligence (AI) and machine learning (ML) in SQLi detection and cybersecurity frameworks. We analyze various ML methodologies, including supervised learning, deep learning (e.g., CNNs, LSTMs), reinforcement learning, and hybrid models, highlighting their effectiveness in identifying sophisticated attack patterns. Additionally, generative AI models such as Variation Auto encoders (VAEs) and Generative Adversarial Networks (GANs) show promise in data augmentation for enhanced detection robustness. Despite their advantages, ML-based cybersecurity solutions face challenges, including dataset imbalance, adversarial ML threats, and computational constraints. The study underscores the necessity of standardized datasets and AI-driven adaptive security frameworks to improve real-time threat detection. Future research should focus on scalable, interpretable AI models, adversarial resilience, and hybrid security approaches to mitigate evolving cyber threats effectively. This review provides valuable insights into the role of AI- driven security mechanisms in combating SQLi attacks and strengthening digital infrastructure protection.