Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Web Technologies Security in the AI Era: A Survey of CDN-Enhanced Defenses

Mehrab Hosain, Sabbir Alom Shuvo, Matthew Ogbe, Md Shah Jalal Mazumder, Yead Rahman, Md Azizul Hakim, Anukul Pandey

TL;DR

The paper surveys CDN-integrated, AI-enabled defenses at the edge (WAAP) for web and API security, arguing that proximity to users improves detection and response while reducing data movement. It introduces a threat taxonomy, edge-signal mapping, and model-design patterns (streaming, drift-aware personalization, calibration, and caching-aware inference), plus deployment playbooks with SLO-centric evaluation. Key contributions include methodological synthesis, edge-specific threat-action mappings, case-driven deployment guidance, and governance considerations to address privacy, adversarial ML, and compliance. The work highlights practical benefits—faster containment and lower data transfer—alongside risks such as model poisoning and governance challenges, and outlines priorities in XAI, robustness, and privacy-preserving learning for edge AI defenses.

Abstract

The modern web stack, which is dominated by browser-based applications and API-first backends, now operates under an adversarial equilibrium where automated, AI-assisted attacks evolve continuously. Content Delivery Networks (CDNs) and edge computing place programmable defenses closest to users and bots, making them natural enforcement points for machine-learning (ML) driven inspection, throttling, and isolation. This survey synthesizes the landscape of AI-enhanced defenses deployed at the edge: (i) anomaly- and behavior-based Web Application Firewalls (WAFs) within broader Web Application and API Protection (WAAP), (ii) adaptive DDoS detection and mitigation, (iii) bot management that resists human-mimicry, and (iv) API discovery, positive security modeling, and encrypted-traffic anomaly analysis. We add a systematic survey method, a threat taxonomy mapped to edge-observable signals, evaluation metrics, deployment playbooks, and governance guidance. We conclude with a research agenda spanning XAI, adversarial robustness, and autonomous multi-agent defense. Our findings indicate that edge-centric AI measurably improves time-to-detect and time-to-mitigate while reducing data movement and enhancing compliance, yet introduces new risks around model abuse, poisoning, and governance.

Web Technologies Security in the AI Era: A Survey of CDN-Enhanced Defenses

TL;DR

The paper surveys CDN-integrated, AI-enabled defenses at the edge (WAAP) for web and API security, arguing that proximity to users improves detection and response while reducing data movement. It introduces a threat taxonomy, edge-signal mapping, and model-design patterns (streaming, drift-aware personalization, calibration, and caching-aware inference), plus deployment playbooks with SLO-centric evaluation. Key contributions include methodological synthesis, edge-specific threat-action mappings, case-driven deployment guidance, and governance considerations to address privacy, adversarial ML, and compliance. The work highlights practical benefits—faster containment and lower data transfer—alongside risks such as model poisoning and governance challenges, and outlines priorities in XAI, robustness, and privacy-preserving learning for edge AI defenses.

Abstract

The modern web stack, which is dominated by browser-based applications and API-first backends, now operates under an adversarial equilibrium where automated, AI-assisted attacks evolve continuously. Content Delivery Networks (CDNs) and edge computing place programmable defenses closest to users and bots, making them natural enforcement points for machine-learning (ML) driven inspection, throttling, and isolation. This survey synthesizes the landscape of AI-enhanced defenses deployed at the edge: (i) anomaly- and behavior-based Web Application Firewalls (WAFs) within broader Web Application and API Protection (WAAP), (ii) adaptive DDoS detection and mitigation, (iii) bot management that resists human-mimicry, and (iv) API discovery, positive security modeling, and encrypted-traffic anomaly analysis. We add a systematic survey method, a threat taxonomy mapped to edge-observable signals, evaluation metrics, deployment playbooks, and governance guidance. We conclude with a research agenda spanning XAI, adversarial robustness, and autonomous multi-agent defense. Our findings indicate that edge-centric AI measurably improves time-to-detect and time-to-mitigate while reducing data movement and enhancing compliance, yet introduces new risks around model abuse, poisoning, and governance.

Paper Structure

This paper contains 30 sections, 3 equations, 5 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Survey Methodology
  3. Foundations and threat Model
  4. Common Web and API Weaknesses
  5. AI/ML Building Blocks
  6. Notation and Anomaly Scoring
  7. System Model and Assumptions
  8. Identity & Session Realities
  9. AI/ML-Powered Defenses
  10. DDoS Detection and Mitigation
  11. Bot Management and Human-Mimicry Resistance
  12. API Security
  13. Threat Taxonomy and Edge-Signal Mapping
  14. Edge Feature and Model Patterns
  15. Streaming Detection Primitives
  16. ...and 15 more sections

Figures (5)

  • Figure 1: AI/ML DDoS pipeline from edge telemetry to adaptive mitigation with feedback.
  • Figure 2: Bot detection flow: behavior + fingerprint features $\rightarrow$ ML classification $\rightarrow$ allow/challenge/block with continuous learning.
  • Figure 3: AI-enhanced CDN path: classification and policy decisions at PoPs prior to origin contact.
  • Figure 4: Federated learning with DP/HE for privacy-preserving edge training and global aggregation.
  • Figure 5: WAAP providers across six AI-enabled dimensions