Detection of Personally Identifiable Information Leakage on the Web Using Artificial Intelligence Techniques: A Comprehensive Review

The rapid digital transformation of services, combined with the growth of web applications, has intensified the threat of personally identifiable information (PII) leakage. Therefore, there is a need for PII leakage detection to maintain the security and privacy of data in the cybersecurity sphere, which has become a research area of interest. This paper aims to review literature on the Personal Data Breach Detection approaches based on web scraping techniques and Artificial Intelligence (AI) tools, including machine learning and deep learning, for automatic detection of PII exposures on different digital platforms. It demonstrates how a hybrid model combining BERT-BiLSTM-Attention and rule-based extraction reaches 99.15% F1 score which reveals its exceptional capability to detect PII leaks. The paper identifies key gaps in current detection techniques and points out the necessity to make real-time automated detection of PII exposure incidents. By consolidating the knowledge about the detection of PII breaches from existing literature, this review gives a comprehensive understanding of how to mitigate detection of PII breaches to help future researchers improve cybersecurity measures.