Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ImmuniFraug: A Metacognitive Intervention Anti-Fraud Approach to Enhance Undergraduate Students' Cyber Fraud Awareness

Xiangzhe Yuan, Jiajun Wang, Huanchen Wang, Qian Wan, Siying Hu

TL;DR

Cyber fraud now dominates crime in China, with undergraduates disproportionately affected. The paper introduces ImmuniFraug, an LLM-based metacognitive intervention that uses immersive, multimodal fraud simulations (text, voice, avatars) across ten fraud types, followed by PMT-informed debriefs to promote reflective learning. In a controlled study with 846 undergraduates, the interactive intervention significantly improved fraud awareness beyond traditional materials (p = 0.026) and achieved high narrative transportation (mean around $M=56.95$ on a 7–77 scale). Qualitative interviews identified realism, adaptive deception, time pressure, emotional manipulation awareness, and self-efficacy as key effectiveness factors, supporting a shift from passive learning to active, metacognitive engagement. Collectively, ImmuniFraug demonstrates a scalable, ecologically valid approach for anti-fraud training with potential to generalize across scam genres and improve long-term resilience against fraud.

Abstract

Cyber fraud now constitutes over half of criminal cases in China, with undergraduate students experiencing a disproportionate rise in victimization. Traditional anti-fraud training remains predominantly passive, yielding limited engagement and retention. This paper introduces ImmuniFraug, a Large Language Model (LLM)-based metacognitive intervention that delivers immersive, multimodal fraud simulations integrating text, voice, and visual avatars across ten prevalent fraud types. Each scenario is designed to replicate real-world persuasion tactics and psychological pressure, while post-interaction debriefs provide grounded feedback in protection motivation theory and reflective prompts to reinforce learning. In a controlled study with 846 Chinese undergraduates, ImmuniFraug was compared to official text-based materials. Linear Mixed-Effects Modeling (LMEM) reveals that the interactive intervention significantly improved fraud awareness (p = 0.026), successfully providing incremental learning value even when controlling for participants' extensive prior exposure to anti-fraud education, alongside high narrative immersion (M = 56.95/77). Thematic analysis of interviews revealed key effectiveness factors: perceived realism, adaptive deception, enforced time pressure, emotional manipulation awareness, and enhanced self-efficacy. Findings demonstrate that by shifting the focus from passive knowledge acquisition to active metacognitive engagement, LLM-based simulations offer a scalable and ecologically valid new paradigm for anti-fraud training and fostering fraud resilience.

ImmuniFraug: A Metacognitive Intervention Anti-Fraud Approach to Enhance Undergraduate Students' Cyber Fraud Awareness

TL;DR

Cyber fraud now dominates crime in China, with undergraduates disproportionately affected. The paper introduces ImmuniFraug, an LLM-based metacognitive intervention that uses immersive, multimodal fraud simulations (text, voice, avatars) across ten fraud types, followed by PMT-informed debriefs to promote reflective learning. In a controlled study with 846 undergraduates, the interactive intervention significantly improved fraud awareness beyond traditional materials (p = 0.026) and achieved high narrative transportation (mean around on a 7–77 scale). Qualitative interviews identified realism, adaptive deception, time pressure, emotional manipulation awareness, and self-efficacy as key effectiveness factors, supporting a shift from passive learning to active, metacognitive engagement. Collectively, ImmuniFraug demonstrates a scalable, ecologically valid approach for anti-fraud training with potential to generalize across scam genres and improve long-term resilience against fraud.

Abstract

Cyber fraud now constitutes over half of criminal cases in China, with undergraduate students experiencing a disproportionate rise in victimization. Traditional anti-fraud training remains predominantly passive, yielding limited engagement and retention. This paper introduces ImmuniFraug, a Large Language Model (LLM)-based metacognitive intervention that delivers immersive, multimodal fraud simulations integrating text, voice, and visual avatars across ten prevalent fraud types. Each scenario is designed to replicate real-world persuasion tactics and psychological pressure, while post-interaction debriefs provide grounded feedback in protection motivation theory and reflective prompts to reinforce learning. In a controlled study with 846 Chinese undergraduates, ImmuniFraug was compared to official text-based materials. Linear Mixed-Effects Modeling (LMEM) reveals that the interactive intervention significantly improved fraud awareness (p = 0.026), successfully providing incremental learning value even when controlling for participants' extensive prior exposure to anti-fraud education, alongside high narrative immersion (M = 56.95/77). Thematic analysis of interviews revealed key effectiveness factors: perceived realism, adaptive deception, enforced time pressure, emotional manipulation awareness, and enhanced self-efficacy. Findings demonstrate that by shifting the focus from passive knowledge acquisition to active metacognitive engagement, LLM-based simulations offer a scalable and ecologically valid new paradigm for anti-fraud training and fostering fraud resilience.
Paper Structure (51 sections, 1 equation, 3 figures, 7 tables)

This paper contains 51 sections, 1 equation, 3 figures, 7 tables.

Table of Contents

  1. Related Work
  2. Background of Cyber Fraud in China
  3. Current Landscape and Limitations of Anti-Fraud Prevention
  4. Metacognitive Intervention Approaches in the Security Domain
  5. Methodology
  6. Design of ImmuniFraug
  7. Process and prompt design
  8. Design of the multimodal interaction
  9. Procedure
  10. Participant Recruitment
  11. Results
  12. RQ1: Effectiveness of LLM-based Simulation in Improving Cyber Fraud Awareness
  13. Preliminary Between-Group Analysis
  14. Effectiveness Evaluation via Linear Mixed-Effects Modeling
  15. RQ2: Perceptions of the Simulation and its Metacognitive Impacts
  16. ...and 36 more sections

Figures (3)

  • Figure 1: Workflow of the ImmuniFraug system and experimental design.
  • Figure 2: User interface and interactive simulation flow in ImmuniFraug. The system integrates text, voice, and avatar-based modalities to deliver realistic fraud scenarios. (Left) User interface examples, where participants engage with avatars via text or voice. (Center) The simulation includes: (a) a pre-simulation disclaimer, (b) interactive fraud dialogue with contextual realism, and (c) branching response options guiding narrative progression. (Right) Sessions terminate when (d.1) the scam is rejected or identified, (d.2) users voluntarily exit, or (d.3) a predefined interaction limit is reached. Finally, (e) a post-simulation debrief provides feedback and reinforces awareness. These design choices balance immersion with controlled exposure, replicating real-world scam dynamics while safeguarding participants, and are grounded in metacognitive principles that foster awareness.
  • Figure 3: Distribution of subjective scores across the four key constructs. The violin plots illustrate the kernel density estimation of the data. Inside, the box plots represent the median (thick line) and interquartile range (IQR), while the overlaid gray jittered points depict the individual perceptual ratings of the 391 participants. Refer to [\ref{['CustomQuestionnaire']}] for the detailed questionnaire items.