The field of web security and privacy is rapidly evolving, with a growing focus on protecting user data and preventing cyber threats. Recent research has highlighted the risks of data leakage in e-commerce platforms, with nearly 30% of popular online shops found to be violating user privacy. Meanwhile, advances in large language models (LLMs) are being explored for their potential in detecting web-based attacks, such as WebShell attacks. However, these models also pose new challenges, such as balancing security and usability in digital gateways. Noteworthy papers in this area include the proposal of a novel classification scheme for web tracking systems, which emphasizes technological mechanisms such as HTTP protocols and user identification techniques. Another notable study demonstrated the effectiveness of an ensemble method consisting of LSTM, GRU, and stacked autoencoders in detecting zero-day web attacks, achieving remarkable detection metrics with an exceptionally low false positive rate. Additionally, a comprehensive review of malicious URL detection techniques, datasets, and code repositories provided valuable insights into the current state of this ongoing battle, highlighting the importance of modal information channels and the need for standardized benchmarking.