Pith. sign in

REVIEW 2 major objections 2 minor 119 references

Generative AI-Enabled Refund Fraud in Chinese E-Commerce: Investigation on Merchants and Platform Workers

T0 review · 2 major / 2 minor · reviewed 2026-06-28 · grok-4.3

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Generative AI invalidates the security assumption that digital evidence reflects physical reality by enabling cheap fabrication of product defect evidence for e-commerce refunds.

desk verdict This interview study maps GenAI refund fraud tactics in Chinese e-commerce via 30 self-reports and offers a four-phase taxonomy, but the core claims lack any external checks or logs. read the letter →

arxiv 2606.03215 v1 pith:26V4IL77 submitted 2026-06-02 cs.CR cs.HC

Shuning Zhang , Eve He , Xiao Zhan , Shijing He , Robert Xiao , Xin Yi , Hewu Li This is my paper
classification cs.CRcs.HC
keywords generativeAIrefundfraude-commercedisputeresolutionthreatvectorsChinesemarketverificationstrategiesmitigation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper shows how generative AI shifts the threat model in e-commerce dispute resolution. Interviews with 17 merchants and 13 platform workers in China reveal attackers synthesizing realistic defect evidence at negligible cost across four phases of transactions. A taxonomy outlines threat vectors in transaction, dispute, logistics, and communication stages. Platforms counter with AI screening and requests for multi-angle videos, yet face structural and technical barriers to effective defense. The work points to privacy-preserving databases and material anchors as ways to restore traceability.

What carries the argument

Taxonomy of four GenAI-enabled threat vectors that let attackers synthesize physically plausible product defects at scale across transaction, dispute, logistics, and communication phases.

What would settle it

A review of actual dispute-resolution records showing no measurable rise in fabricated defect evidence after widespread GenAI availability would falsify the central claim.

Watch

Extended reading notes

Core claim

Generative AI invalidates this threat model, enabling attackers to fabricate hyper-realistic evidence of product defects at negligible cost. Through semi-structured interviews with merchants (N=17) and platform workers (N=13) in the Chinese e-commerce market, we characterize this shift toward GenAI-enabled scalable fabrication. We outline a taxonomy of four GenAI-enabled threat vectors across the transaction, dispute, logistics and communication phases, highlighting how attackers exploit GenAI to synthesize physically plausible product defects at scale.

Load-bearing premise

The semi-structured interview responses from the 30 participants accurately capture the prevalence, methods, and evolution of GenAI-enabled refund fraud practices in the Chinese market.

Editorial extensions

If this is right

  • Platforms and merchants adapt verification strategies by relying on AI tools for automated screening and adversarial interrogation such as requesting multi-angle videos.
  • Adoption of these defenses faces implementation hurdles like structural platform constraints and fundamental limitations regarding the technical sophistication of GenAI.
  • Design implications include privacy-preserving cross-platform fraud databases.
  • Traceability mechanisms such as embedding verifiable material anchors into the product can help restore the link between digital evidence and physical reality.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same GenAI fabrication techniques could spread to e-commerce platforms outside China once tools become widely available.
  • Heightened verification demands may increase operational costs or friction for honest merchants and buyers.
  • Material-anchor approaches would require coordination across supply chains to embed verifiable markers at production time.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 2 minor

Summary. The manuscript claims that generative AI invalidates traditional e-commerce dispute-resolution threat models by enabling attackers to fabricate hyper-realistic evidence of product defects at negligible cost. Drawing on semi-structured interviews with 17 merchants and 13 platform workers in the Chinese market, the authors present a taxonomy of four GenAI-enabled threat vectors spanning the transaction, dispute, logistics, and communication phases; describe merchant and platform adaptations such as AI screening and requests for multi-angle videos; identify adoption challenges; and outline design implications including privacy-preserving fraud databases and material traceability anchors.

Significance. If the reported practices prove accurate and widespread, the work identifies a material evolution in e-commerce fraud that could affect dispute-resolution design and platform policy. The taxonomy supplies a concrete organizing framework for subsequent technical and empirical studies, and the adaptation challenges highlight actionable tensions between verification strength and operational constraints. The absence of quantitative prevalence data or technical validation, however, confines the contribution to an exploratory characterization rather than a definitive demonstration of invalidated threat models.

major comments (2)
  1. [Methods] Methods section: the description of the 30 semi-structured interviews supplies no information on recruitment, sampling frame, interview protocol, transcription, coding process, or any form of validation or inter-rater reliability. Because the four-vector taxonomy and all claims about scalable, negligible-cost fabrication rest exclusively on themes extracted from these self-reports, the missing methodological detail is load-bearing for the central empirical contribution.
  2. [Findings / Threat Vectors] Findings on threat vectors (§4 or equivalent): the assertions that GenAI enables 'hyper-realistic' defect fabrication 'at negligible cost' and thereby 'invalidates' the evidence-based threat model are presented as established facts derived from interviewee statements, yet the manuscript reports neither controlled tests of GenAI output realism, cost measurements, nor triangulation against platform dispute logs or forensic case data. This evidentiary gap directly affects the strength of the invalidation claim.
minor comments (2)
  1. [Abstract] The abstract states N=17 merchants and N=13 platform workers; the main text should explicitly confirm whether any participants held dual roles or whether the samples are fully disjoint.
  2. [Discussion / Design Implications] The design-implications section would benefit from a short paragraph acknowledging that the proposed cross-platform database and material-anchor mechanisms remain high-level and have not been prototyped or evaluated within the study.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive and detailed feedback, which highlights important areas for strengthening the manuscript's transparency and scope. We address each major comment below, indicating where revisions will be made to improve methodological detail and clarify the nature of our claims while preserving the exploratory qualitative contribution.

read point-by-point responses

  1. Referee: [Methods] Methods section: the description of the 30 semi-structured interviews supplies no information on recruitment, sampling frame, interview protocol, transcription, coding process, or any form of validation or inter-rater reliability. Because the four-vector taxonomy and all claims about scalable, negligible-cost fabrication rest exclusively on themes extracted from these self-reports, the missing methodological detail is load-bearing for the central empirical contribution.

    Authors: We agree that the current Methods section is insufficiently detailed and that this information is essential for assessing the taxonomy's foundation. In the revised manuscript we will expand the section to specify: recruitment through purposive outreach via Chinese e-commerce professional networks and targeted online communities; sampling frame using purposive selection for diversity in merchant scale and platform worker experience levels; interview protocol as a semi-structured guide with core questions on observed GenAI fraud patterns, evidence fabrication, and platform responses; transcription as verbatim from audio recordings with subsequent translation; coding via iterative thematic analysis performed independently by two researchers on an initial subset of transcripts followed by consensus discussion; and validation steps including member checking with a subset of participants. These additions will make the empirical grounding explicit. revision: yes

  2. Referee: [Findings / Threat Vectors] Findings on threat vectors (§4 or equivalent): the assertions that GenAI enables 'hyper-realistic' defect fabrication 'at negligible cost' and thereby 'invalidates' the evidence-based threat model are presented as established facts derived from interviewee statements, yet the manuscript reports neither controlled tests of GenAI output realism, cost measurements, nor triangulation against platform dispute logs or forensic case data. This evidentiary gap directly affects the strength of the invalidation claim.

    Authors: The manuscript is framed as an exploratory qualitative study of stakeholder-reported practices rather than a technical validation of GenAI capabilities. The statements on hyper-realistic fabrication and negligible cost reflect consistent themes reported by the 17 merchants and 13 platform workers based on their direct encounters. We do not present these as independently verified technical facts. To address the concern we will revise the abstract, introduction, and §4 to use phrasing such as 'interviewees report that GenAI enables...' and 'this is perceived to shift the practical threat model,' add an explicit Limitations section noting the lack of controlled tests, cost measurements, quantitative prevalence data, and external triangulation, and position the taxonomy as an organizing framework to guide future empirical work. This maintains the contribution's scope while clarifying its evidentiary basis. revision: partial

standing simulated objections not resolved
  • Providing controlled tests of GenAI output realism, direct cost measurements, or triangulation with platform dispute logs and forensic data, as these require experimental designs and data access outside the scope of this qualitative interview study.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: empirical interview study with no derivations or self-referential modeling

full rationale

The paper reports themes extracted from semi-structured interviews (N=17 merchants, N=13 platform workers) to characterize GenAI-enabled refund fraud. No equations, fitted parameters, quantitative predictions, or derivation chains exist. Claims rest directly on participant responses without any self-citation load-bearing steps, uniqueness theorems, ansatzes, or renamings of known results. The taxonomy and adaptation challenges are presented as inductive findings from the data rather than outputs forced by prior inputs or definitions. This is a standard qualitative investigation whose central assertions are independent of any circular reduction.

Assumptions & free parameters 0 free parameters · 1 assumptions · 0 invented entities

The central claim rests on the validity and representativeness of qualitative interview data from a single market; no free parameters or invented entities are present.

assumptions (1)
  • domain assumption Semi-structured interviews with merchants and platform workers provide reliable insights into actual GenAI fraud practices.
    The taxonomy and threat vectors are built directly from these self-reported experiences.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generative AI-Enabled Refund Fraud in Chinese E-Commerce: Investigation on Merchants and Platform Workers." pith.science (2026). https://pith.science/paper/26V4IL77

@misc{pith2026260603215,
  author       = {Pith},
  title        = {Pith review of: Generative AI-Enabled Refund Fraud in Chinese E-Commerce: Investigation on Merchants and Platform Workers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/26V4IL77}},
  note         = {Machine review of arXiv:2606.03215}
}
read the original abstract

E-commerce dispute resolution typically relies on the security assumption that digital evidence truthfully reflects physical reality. Generative AI (GenAI) invalidates this threat model, enabling attackers to fabricate hyper-realistic evidence of product defects at negligible cost. Through semi-structured interviews with merchants (N=17) and platform workers (N=13) in the Chinese e-commerce market, we characterize this shift toward GenAI-enabled scalable fabrication. We outline a taxonomy of four GenAI-enabled threat vectors across the transaction, dispute, logistics and communication phases, highlighting how attackers exploit GenAI to synthesize physically plausible product defects at scale. To mitigate these threats, platforms and merchants are adapting verification strategies, relying on AI tools for automated screening and adversarial interrogation (e.g., requesting multi-angle videos) to increase attack complexity. However, we find several challenges that hinder the adoption of these defenses, including implementation hurdles like structural platform constraints and fundamental limitations regarding the technical sophistication of GenAI. We conclude by outlining design implications for privacy-preserving cross-platform fraud databases, and traceability mechanisms such as embedding verifiable material anchors into the product.

Figures

Figures reproduced from arXiv: 2606.03215 by the authors.

Figure 1
Figure 1. The ecosystem around refund fraud, involving cus [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Illustrations of manipulated images used for refund [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

119 extracted references · 3 canonical work pages

  1. [1]

    Zainul Abi Din, Hari Venugopalan, Jaime Park, Andy Li, Weisu Yin, HaoHui Mai, Yong Jae Lee, Steven Liu, and Samuel T King. 2020. Boxer: Preventing fraud by scanning credit cards. In29th USENIX Security Symposium (USENIX Security 20). 1571–1588

    2020

  2. [2]

    Alibaba Group. 2025. Taobao. https://www.taobao.com/. [Accessed: 2026-04-29]

    2025

  3. [3]

    2025.Inside Amazon’s fight against organized refund fraud

    Amazon Staff. 2025.Inside Amazon’s fight against organized refund fraud. https://trustworthyshopping.aboutamazon.com/inside-amazons-fight- against-refund-fraud [Accessed: 2026-04-29]

    2025

  4. [4]

    Daniel Arp, Erwin Quiring, Tammo Krueger, Stanimir Dragiev, and Konrad Rieck

  5. [5]

    InInternational Conference on Security and Privacy in Communication Systems

    Privacy-enhanced fraud detection with Bloom filters. InInternational Conference on Security and Privacy in Communication Systems. Springer, 396– 415

  6. [6]

    Arjun Arunasalam and Z Berkay Celik. 2025. Characterizing e-commerce harm by investigating online communities: A case study with abusive dropshipping. IEEE Security & Privacy(2025)

    2025

  7. [7]

    Arjun Arunasalam, Andrew Chu, Muslum Ozgur Ozmen, Habiba Farrukh, and Z Berkay Celik. 2024. The Dark Side of E-Commerce: Dropshipping Abuse as a Business Model.. InNDSS

    2024

  8. [8]

    Sumit Badotra, Amit Sundas, et al . 2021. A systematic review on security of E-commerce systems.International Journal of Applied Science and Engineering 18, 2 (2021), 1–19

    2021

Show all 119 references
  1. [9]

    Michael Bailey, David Dittrich, Erin Kenneally, and Doug Maughan. 2012. The menlo report.IEEE Security & Privacy10, 2 (2012), 71–75

    2012

  2. [10]

    Tom L Beauchamp et al. 2008. The belmont report.The Oxford textbook of clinical research ethics(2008), 149–155

    2008

  3. [11]

    Marzieh Bitaab, Haehyun Cho, Adam Oest, Zhuoer Lyu, Wei Wang, Jorij Abraham, Ruoyu Wang, Tiffany Bao, Yan Shoshitaishvili, and Adam Doupé. 2023. Beyond 7https://pages.ravelin.com/agentic-commerce-fraud-report/ phish: Toward detecting fraudulent e-commerce websites at scale. In...

    2023

  4. [12]

    Virginia Braun and Victoria Clarke. 2006. Using thematic analysis in psychology. Qualitative research in psychology3, 2 (2006), 77–101

    2006

  5. [13]

    ByteDance. 2025. Douyin E-commerce. https://douyinec.com/. [Accessed: 2026-04-29]

    2025

  6. [14]

    Claudio Carpineto and Giovanni Romano. 2017. Learning to detect and mea- sure fake ecommerce websites in search-engine results. InProceedings of the international conference on web intelligence. 403–410

    2017

  7. [15]

    Neha Chachra, Stefan Savage, and Geoffrey M Voelker. 2015. Affiliate crook- ies: Characterizing affiliate marketing abuse. InProceedings of the 2015 Internet Measurement Conference. 41–47

    2015

  8. [16]

    Rujia Chen, Yi Xie, Minglang Liao, Jiankun Hu, and Xingcheng Liu. 2025. Iden- tifying Fraudulent Users in E-commerce Applications through Spatiotemporal Fusion and Selective Aggregation.ACM Transactions on Privacy and Security29, 1 (2025), 1–27

    2025

  9. [17]

    Beomsang Cho, Binh M Le, Jiwon Kim, Simon Woo, Shahroz Tariq, Alsharif Abuadbba, and Kristen Moore. 2023. Towards understanding of deepfake videos in the wild. InProceedings of the 32nd ACM international conference on information and knowledge management. 4530–4537

    2023

  10. [18]

    Joe DeBlasio, Saikat Guha, Geoffrey M Voelker, and Alex C Snoeren. 2017. Explor- ing the dynamics of search advertiser fraud. InProceedings of the 2017 Internet Measurement Conference. 157–170

    2017

  11. [19]

    I Cannot Write This Because It Violates Our Content Policy

    Lan Gao, Oscar Chen, Rachel Lee, Nick Feamster, Chenhao Tan, and Marshini Chetty. 2025. " I Cannot Write This Because It Violates Our Content Policy": Understanding Content Moderation Policies and User Experiences in Generative {AI} Products. In34th USENIX Security Symposium (...

    2025

  12. [20]

    Jamie George. 2024. AI-powered refund abuse and dispute fraud: The democrati- zation of deception. Ravelin Blog. https://www.ravelin.com/blog/ai-powered- refund-abuse-dispute-fraud [Accessed: 2026-04-29]

    2024

  13. [21]

    Zhihao Gu, Yang Chen, Taiping Yao, Shouhong Ding, Jilin Li, Feiyue Huang, and Lizhuang Ma. 2021. Spatiotemporal inconsistency learning for deepfake video detection. InProceedings of the 29th ACM international conference on multimedia. 3473–3481

    2021

  14. [22]

    Desta Haileselassie Hagos, Rick Battle, and Danda B Rawat. 2024. Recent advances in generative ai and large language models: Current status, challenges, and perspectives.IEEE transactions on artificial intelligence(2024)

    2024

  15. [23]

    Shuang Hao, Kevin Borgolte, Nick Nikiforakis, Gianluca Stringhini, Manuel Egele, Michael Eubanks, Brian Krebs, and Giovanni Vigna. 2015. Drops for stuff: An analysis of reshipping mule scams. InProceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Securi...

    2015

  16. [24]

    Yongquan ‘Owen’ Hu, Jingyu Tang, Xinya Gong, Zhongyi Zhou, Shuning Zhang, Don Samitha Elvitigala, Florian ‘Floyd’ Mueller, Wen Hu, and Aaron J Quigley

  17. [25]

    InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems

    Vision-based multimodal interfaces: A survey and taxonomy for enhanced context-aware system design. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. 1–31

    2025

  18. [26]

    JD.com, Inc. 2025. JD.com. https://www.jd.com/. [Accessed: 2026-04-29]

    2025

  19. [27]

    Man Ji, Xiayu Chen, and Shaobo Wei. 2025. What motivates consumers’ pur- chase intentions in e-commerce live streaming: A socio-technical perspective. International Journal of Human–Computer Interaction41, 2 (2025), 1585–1605

    2025

  20. [28]

    Matthew Joslin, Xian Wang, and Shuang Hao. 2024. Double Face: Leveraging User Intelligence to Characterize and Recognize {AI-synthesized} Faces. In33rd USENIX Security Symposium (USENIX Security 24). 1009–1026

    2024

  21. [29]

    Prakash L Kharvi. 2024. Understanding the impact of AI-generated deepfakes on public opinion, political discourse, and personal security in social media.IEEE Security & Privacy22, 4 (2024), 115–122

    2024

  22. [30]

    Maxim Kolomeets, Han Wu, Lei Shi, and Aad Van Moorsel. 2024. The face of deception: The impact of AI-generated photos on malicious social bots.IEEE Transactions on Computational Social Systems12, 3 (2024), 1080–1091

    2024

  23. [31]

    Platon Kotzias, Michalis Pachilakis, Javier Aldana-Iuit, Juan Caballero, Iskander Sánchez-Rola, and Leyla Bilge. 2025. Ctrl+ Alt+ Deceive: Quantifying User Exposure to Online Scams.. InNDSS

    2025

  24. [32]

    Platon Kotzias, Kevin Roundy, Michalis Pachilakis, Iskander Sanchez-Rola, and Leyla Bilge. 2023. Scamdog millionaire: Detecting e-commerce scams in the wild. InProceedings of the 39th Annual Computer Security Applications Conference. 29–43

    2023

  25. [33]

    Gali Ramesh Kumar and N Umadevi. 2024. A Multi-Perspective Fraud Detection Method for Multi-Participant E-Commerce Transactions.International Journal of Mechanical Engineering Research and Technology16, 2 (2024), 350–364

    2024

  26. [34]

    Srijan Kumar, Bryan Hooi, Disha Makhija, Mohit Kumar, Christos Faloutsos, and VS Subrahmanian. 2018. Rev2: Fraudulent user prediction in rating platforms. In Proceedings of the eleventh ACM international conference on web search and data mining. 333–341

    2018

  27. [35]

    Yuezun Li, Cong Zhang, Pu Sun, Lipeng Ke, Yan Ju, Honggang Qi, and Siwei Lyu

  28. [36]

    In2021 IEEE Security and Privacy Workshops (SPW)

    DeepFake-o-meter: An open platform for DeepFake detection. In2021 IEEE Security and Privacy Workshops (SPW). IEEE, 277–281. Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Trovato et al

    2018

  29. [37]

    Xiao Ling, David Yan, Bilal Alsallakh, Ashutosh Pandey, Manan Bakshi, and Pamela Bhattacharya. 2023. Learned temporal aggregations for fraud classifi- cation on e-commerce platforms. InCompanion Proceedings of the ACM Web Conference 2023. 1365–1372

    2023

  30. [38]

    Mingxuan Liu, Yunyi Zhang, Lijie Wu, Baojun Liu, Geng Hong, Yiming Zhang, Hui Jiang, Jia Zhang, Haixin Duan, Min Zhang, et al . 2025. {NOKEScam}: Understanding and Rectifying {Non-Sense} Keywords Spear Scam in Search Engines. In34th USENIX Security Symposium (USENIX Security 2...

    2025

  31. [39]

    Zeyu Lu, Di Huang, Lei Bai, Jingjing Qu, Chengyue Wu, Xihui Liu, and Wanli Ouyang. 2023. Seeing is not always believing: Benchmarking human and model perception of ai-generated images.Advances in neural information processing systems36 (2023), 25435–25447

    2023

  32. [40]

    Wenrui Ma, Ying Cong, Haitao Xu, Fan Zhang, Zhao Li, and Siqi Ren. 2023. Investi- gating Fraud and Misconduct in Legitimate Internet Economy based on Customer Complaints. In2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (...

    2023

  33. [41]

    Francisco J Martínez-López, Changyuan Feng, Yangchun Li, and David López- López. 2022. Using instant refunds to improve online return experiences.Journal of Retailing and Consumer Services68 (2022), 103067

    2022

  34. [42]

    Iren Mazloomzadeh, Gias Uddin, Foutse Khomh, and Ashkan Sami. 2024. Repu- tation Gaming in Crowd Technical Knowledge Sharing.ACM Transactions on Software Engineering and Methodology34, 1 (2024), 1–41

    2024

  35. [43]

    Nora McDonald, Sarita Schoenebeck, and Andrea Forte. 2019. Reliability and inter-rater reliability in qualitative research: Norms and guidelines for CSCW and HCI practice.Proceedings of the ACM on human-computer interaction3, CSCW (2019), 1–23

    2019

  36. [44]

    Jaron Mink, Licheng Luo, Natã M Barbosa, Olivia Figueira, Yang Wang, and Gang Wang. 2022. {DeepPhish}: Understanding user trust towards artificially generated profiles in online social networks. In31st USENIX Security Symposium (USENIX Security 22). 1669–1686

    2022

  37. [45]

    Jaron Mink, Miranda Wei, Collins W Munyendo, Kurt Hugenberg, Tadayoshi Kohno, Elissa M Redmiles, and Gang Wang. 2024. It’s trying too hard to look real: deepfake moderation mistakes and identity-based bias. InProceedings of the 2024 CHI Conference on Human Factors in Computing...

    2024

  38. [46]

    Abed Mutemi and Fernando Bacao. 2024. E-commerce fraud detection based on machine learning techniques: Systematic literature review.Big Data Mining and Analytics7, 2 (2024), 419–444

    2024

  39. [47]

    Asmit Nayak, Yash Wani, Shirley Zhang, Rishabh Khandelwal, and Kassem Fawaz

  40. [48]

    InProceedings of the 2025 ACM SIGSAC Conference on Computer and Communications Security

    Automatically Detecting Online Deceptive Patterns. InProceedings of the 2025 ACM SIGSAC Conference on Computer and Communications Security. 96–110

    2025

  41. [49]

    Rajvardhan Oak and Zubair Shafiq. 2025. Towards Characterizing and Detecting Incentivized Reviews on eCommerce Platforms. InProceedings of the International AAAI Conference on Web and Social Media, Vol. 19. 1358–1369

    2025

  42. [50]

    Cameron O’Sullivan. 2025. The impact of e-commerce fraud in reverse logistics. (2025)

    2025

  43. [51]

    Rohan Pagey, Mohammad Mannan, and Amr Youssef. 2023. All your shops are belong to us: security weaknesses in e-commerce platforms. InProceedings of the ACM Web Conference 2023. 2144–2154

    2023

  44. [52]

    Peter S Park, Simon Goldstein, Aidan O’Gara, Michael Chen, and Dan Hendrycks

  45. [53]

    AI deception: A survey of examples, risks, and potential solutions.Patterns 5, 5 (2024)

    2024

  46. [54]

    Mary Anne Patton and Audun Jøsang. 2004. Technologies for trust in electronic commerce.Electronic Commerce Research4, 1 (2004), 9–21

    2004

  47. [55]

    PDD Holdings Inc. 2025. Pinduoduo. https://www.pinduoduo.com/. [Accessed: 2026-04-29]

    2025

  48. [56]

    Christian Perry. 2026. 6 Common Types Of Image-Based Fraud In Refund Work- flows. TruthScan Research Blog. https://truthscan.com/blog/types-of-image- based-fraud/ [Accessed: 2026-04-29]

    2026

  49. [57]

    Saeedreza Shehnepoor, Roberto Togneri, Wei Liu, and Mohammed Bennamoun

  50. [58]

    DFraud 3: multi-component fraud detection free of cold-start.IEEE Trans- actions on Information Forensics and Security16 (2021), 3456–3468

    2021

  51. [59]

    Peter Snyder and Chris Kanich. 2016. Characterizing fraud and its ramifications in affiliate marketing networks.Journal of Cybersecurity2, 1 (2016), 71–81

    2016

  52. [60]

    Kevin Springborn and Paul Barford. 2013. Impression fraud in on-line advertising via {Pay-Per-View} networks. In22nd USENIX Security Symposium (USENIX Security 13). 211–226

    2013

  53. [61]

    Suibin Sun, Le Yu, Xiaokuan Zhang, Minhui Xue, Ren Zhou, Haojin Zhu, Shuang Hao, and Xiaodong Lin. 2021. Understanding and detecting mobile ad fraud through the lens of invalid traffic. InProceedings of the 2021 ACM SIGSAC Confer- ence on Computer and Communications Security. 287–303

    2021

  54. [62]

    Zhibo Sun, Adam Oest, Penghui Zhang, Carlos Rubio-Medrano, Tiffany Bao, Ruoyu Wang, Ziming Zhao, Yan Shoshitaishvili, Adam Doupé, and Gail-Joon Ahn

  55. [63]

    In30th USENIX Security Symposium (USENIX Security 21)

    Having Your Cake and Eating It: An Analysis of {Concession-Abuse-as-a- Service}. In30th USENIX Security Symposium (USENIX Security 21). 4169–4186

  56. [64]

    Li Tang, Qingqing Ye, Haibo Hu, Qiao Xue, Yaxin Xiao, and Jin Li. 2024. DeepMark: A scalable and robust framework for DeepFake video detection.ACM Transactions on Privacy and Security27, 1 (2024), 1–26

    2024

  57. [65]

    Eleanor Tursman, Marilyn George, Seny Kamara, and James Tompkin. 2020. To- wards untrusted social video verification to combat deepfakes via face geometry consistency. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops. 654–655

    2020

  58. [66]

    Rashi Tyagi, Gauri Goyal, and Shiva Tyagi. 2025. Generative AI in Real-Time E-Commerce Fraud Detection: A Comparative and Ethical Analysis. In2025 International Conference on Intelligent Control, Computing and Communications (IC3). IEEE, 269–274

    2025

  59. [67]

    Rolf Van Wegberg, Samaneh Tajalizadehkhoob, Kyle Soska, Ugur Akyazi, Car- los Hernandez Ganan, Bram Klievink, Nicolas Christin, and Michel Van Eeten

  60. [68]

    In27th USENIX security symposium (USENIX security 18)

    Plug and prey? measuring the commoditization of cybercrime via online anonymous markets. In27th USENIX security symposium (USENIX security 18). 1009–1026

  61. [69]

    Bimal Viswanath, M Ahmad Bashir, Mark Crovella, Saikat Guha, Krishna P Gummadi, Balachander Krishnamurthy, and Alan Mislove. 2014. Towards detect- ing anomalous user behavior in online social networks. In23rd usenix security symposium (usenix security 14). 223–238

    2014

  62. [70]

    Cheng Wang and Hangyu Zhu. 2020. Representing fine-grained co-occurrences for behavior-based fraud detection in online payment services.IEEE transactions on dependable and secure computing19, 1 (2020), 301–315

    2020

  63. [71]

    Peng Wang Wang, Xiaojing Liao Liao, Yue Qin, and XiaoFeng Wang. 2020. Into the deep web: Understanding e-commercefraud from autonomous chat with cybercriminals. InProceedings of the ISOC Network and Distributed System Security Symposium (NDSS), 2020

    2020

  64. [72]

    Ye Wang, Zhicong Lu, Peng Cao, Jingyi Chu, Haonan Wang, and Roger Watten- hofer. 2022. How live streaming changes shopping decisions in E-commerce: A study of live streaming commerce.Computer Supported Cooperative Work (CSCW) 31, 4 (2022), 701–729

    2022

  65. [73]

    Kevin Warren, Tyler Tucker, Anna Crowder, Daniel Olszewski, Allison Lu, Caro- line Fedele, Magdalena Pasternak, Seth Layton, Kevin Butler, Carrie Gates, et al

  66. [74]

    Better Be Computer or I’m Dumb

    " Better Be Computer or I’m Dumb": A Large-Scale Evaluation of Humans as Audio Deepfake Detectors. InProceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security. 2696–2710

    2024

  67. [75]

    Elliott Wen, Jiannong Cao, Jiaxing Shen, and Xuefeng Liu. 2018. Fraus: Launching cost-efficient and scalable mobile click fraud has never been so easy. In2018 IEEE Conference on Communications and Network Security (CNS). IEEE, 1–9

    2018

  68. [76]

    Haiqin Weng, Shouling Ji, Fuzheng Duan, Zhao Li, Jianhai Chen, Qinming He, and Ting Wang. 2019. Cats: cross-platform e-commerce fraud detection. In2019 ieee 35th international conference on data engineering (icde). IEEE, 1874–1885

    2019

  69. [77]

    Ancheng Xu, Zhihao Yang, Jingpeng Li, Guanghu Yuan, Longze Chen, Liang Yan, Jiehui Zhou, Zhen Qin, Hengyun Chang, Hamid Alinejad-Rokny, et al. 2025. EVADE: Multimodal Benchmark for Evasive Content Detection in E-Commerce Applications.arXiv preprint arXiv:2505.17654(2025)

    2025 arXiv

  70. [78]

    Chang Xu, Jie Zhang, and Zhu Sun. 2017. Online Reputation Fraud Campaign Detection in User Ratings.. InIJCAI. 3873–3879

    2017

  71. [79]

    Haitao Xu, Daiping Liu, Haining Wang, and Angelos Stavrou. 2015. E-commerce reputation manipulation: The emergence of reputation-escalation-as-a-service. In Proceedings of the 24th International Conference on World Wide Web. 1296–1306

    2015

  72. [80]

    Haitao Xu, Daiping Liu, Haining Wang, and Angelos Stavrou. 2017. An empirical investigation of ecommerce-reputation-escalation-as-a-service.ACM Transac- tions on the Web (TWEB)11, 2 (2017), 1–35

    2017

  73. [81]

    2025.{VoiceWukong}: Benchmarking Deepfake Voice Detection

    Ziwei Yan, Yanjie Zhao, and Haoyu Wang. 2025.{VoiceWukong}: Benchmarking Deepfake Voice Detection. In34th USENIX Security Symposium (USENIX Security 25). 4561–4580

    2025

  74. [82]

    Jianke Yu, Hanchen Wang, Xiaoyang Wang, Zhao Li, Lu Qin, Wenjie Zhang, Jian Liao, Ying Zhang, and Bailin Yang. 2024. Temporal insights for group-based fraud detection on e-commerce platforms.IEEE Transactions on Knowledge and Data Engineering37, 2 (2024), 951–965

    2024

  75. [83]

    Bolin Zhang, Qilin Yin, Wei Lu, and Xiangyang Luo. 2024. Deepfake detection and localization using multi-view inconsistency measurement.IEEE Transactions on Dependable and Secure Computing22, 2 (2024), 1796–1809

    2024

  76. [84]

    Chuang Zhang, Qizhou Wang, Tengfei Liu, Xun Lu, Jin Hong, Bo Han, and Chen Gong. 2021. Fraud detection under multi-sourced extremely noisy annotations. In Proceedings of the 30th ACM international conference on information & knowledge management. 2497–2506

    2021

  77. [85]

    Ge Zhang, Zhao Li, Jiaming Huang, Jia Wu, Chuan Zhou, Jian Yang, and Jianliang Gao. 2022. efraudcom: An e-commerce fraud detection system via competitive graph neural networks.ACM Transactions on Information Systems (TOIS)40, 3 (2022), 1–29

    2022

  78. [86]

    Confrontation or Acceptance

    Shuning Zhang and Shixuan Li. 2024. " Confrontation or Acceptance": Under- standing Novice Visual Artists’ Perception towards AI-assisted Art Creation. arXiv preprint arXiv:2410.14925(2024)

    2024

  79. [87]

    Shuning Zhang, Hui Wang, and Xin Yi. 2025. Exploring collaboration patterns and strategies in human-ai co-creation through the lens of agency: A scoping review of the top-tier hci literature.Proceedings of the ACM on Human-Computer Interaction9, 7 (2025), 1–43

    2025

  80. [88]

    Shuning Zhang, Xin Yi, Haobin Xing, Lyumanshan Ye, Yongquan Hu, and Hewu Li. 2024. Adanonymizer: interactively navigating and balancing the duality Generative AI-Enabled Refund Fraud in Chinese E-Commerce: Investigation on Merchants and Platform Workers Conference acronym ’XX,...

    2024

  81. [89]

    Wenhua Zhang, Weicheng Li, Xuanrong Rao, Lixin Zou, Xiangyang Luo, Chubin Zhuang, Yongjie Hong, Zhen Qin, Hengyu Chang, Chenliang Li, et al . 2025. AIGuard: A Benchmark and Lightweight Detection for E-commerce AIGC Risks. InFindings of the Association for Computational Linguis...

    2025

  82. [90]

    Xiaohan Zhang, Haoqi Ye, Ziqi Huang, Xiao Ye, Yinzhi Cao, Yuan Zhang, and Min Yang. 2023. Understanding the (in) security of cross-side face verification systems in mobile apps: A system perspective. In2023 IEEE Symposium on Security and Privacy (SP). IEEE, 934–950

    2023

  83. [91]

    Yu Zhang, Jing Bian, and Weixiang Zhu. 2013. Trust fraud: A crucial challenge for China’s e-commerce market.Electronic Commerce Research and Applications 12, 5 (2013), 299–308

    2013

  84. [92]

    Yuexin Zhao, Siyi Tang, Hongyu Zhang, and Long Lyu. 2025. AI vs. human: A large-scale analysis of AI-generated fake reviews, human-generated fake reviews and authentic reviews.Journal of Retailing and Consumer Services87 (2025), 104400

    2025

  85. [93]

    Tong Zhu, Yan Meng, Haotian Hu, Xiaokuan Zhang, Minhui Xue, and Haojin Zhu. 2021. Dissecting click fraud autonomy in the wild. InProceedings of the 2021 ACM SIGSAC Conference on Computer and Communications Security. 271–286

    2021

  86. [94]

    Wenjun Zhu, Xiaoyu Ji, Xinfeng Li, Qihang Chen, Kun Wang, Xinyu Li, Ruoyan Xu, and Wenyuan Xu. 2025. Neural Invisibility Cloak: Concealing Adversary in Images via Compromised {AI-driven} Image Signal Processing. In34th USENIX Security Symposium (USENIX Security 25). 937–956

    2025

  87. [95]

    Yongchun Zhu, Dongbo Xi, Bowen Song, Fuzhen Zhuang, Shuai Chen, Xi Gu, and Qing He. 2020. Modeling users’ behavior sequences with hierarchical explainable network for cross-domain fraud detection. InProceedings of the web conference

    2020

  88. [96]

    A Ethical Considerations We acknowledged potential ethical concerns and considered these implications throughout the interviews

    928–938. A Ethical Considerations We acknowledged potential ethical concerns and considered these implications throughout the interviews. Our research was con- ducted with the approval of our institution’s Institutional Review Board (IRB). We detailed our considerations, in ac...

  89. [97]

    Could you briefly describe your role, your experience level, and the types of disputes you typically handle?

  90. [98]

    What are the primary performance metrics or key performance indicators (KPIs) that guide your daily workflow? Part 2: GenAI threat landscape and encounters

  91. [99]

    Can you describe a recent dispute where you suspected a buyer was using AI or automated software, whether to generate a fake persona, fabricate evidence, or manipulate the platform’s logistics?

  92. [100]

    Can you walk me through a case where you strongly suspected the evidence was fabricated, but you ultimately could not definitely prove it or had to approve the refund anyway? What made the evidence difficult to verify? Part 3: Verification practices

  93. [101]

    When you suspect fraudulent evidence, what is your standard verification process?

  94. [102]

    How do you interact with the platform’s risk-screening tools and resolution process?

  95. [103]

    Can you share an instance where you felt the platform’s au- tomated systems misclassified a case? How did you handle that discrepancy?

  96. [104]

    (If applicable) What are your knowledge source for these verification or tools used? Part 4: Challenges

  97. [105]

    In your experience, what are the primary challenges you face when encountering fraudulent activities?

  98. [106]

    Are there rules that you feel constrain your ability to investi- gate refund fraud fraudsters? Why? Part 5: Countermeasures

  99. [107]

    What specific tools, shared databases, or verification stan- dards would help your dispute resolution process? Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Trovato et al

    2018

  100. [108]

    Professional background and general workflow

    Is there anything else you would like to share about around refund fraud? B.2 Interview Script for Merchants Part 1. Professional background and general workflow

  101. [109]

    Could you briefly describe your experience managing your store and your general professional background in e-commerce?

  102. [110]

    Specific cases of synthetic evidence

    What is your standard workflow when you receive and process refund requests from customers? Part 2. Specific cases of synthetic evidence

  103. [111]

    Can you share a recent case where you felt a buyer submitted AI-edited or synthetically altered evidence to justify a refund?

  104. [112]

    What was your initial reaction upon seeing the request and the provided evidence?

  105. [113]

    How did you go about identifying the fraud? What specific details, anomalies, or tools did you rely upon to determine the evidence was likely synthetic?

  106. [114]

    Decision-making logic and operational practices

    Could you walk me through the reasoning process that led you to that conclusion? Part 3. Decision-making logic and operational practices

  107. [115]

    In situations where you suspect the evidence is synthetic, what is your logic for deciding whether to immediately issue the refund, or to reject the request and initiate an appeal?

  108. [116]

    Challenges and expectations

    How do you balance these decisions within your daily opera- tional practices, considering different factors? Part 4. Challenges and expectations

  109. [117]

    What are the primary challenges or limitations you encounter?

  110. [118]

    What specific countermeasures do you expect or hope to see developed in the future to protect merchants?

  111. [119]

    Price” column. Undisclosed criteria are denoted by “–

    Is there anything else you would like to add regarding your experiences and refund fraud? C Participants’ Demographics Table 4 and 5 showed the merchants and platform workers’ demo- graphics separately. Table 4: Participants’ demographics (merchants group). ID Category Platfor...

    2018

Pith tools

Reviewed June 28, 2026 · model on record in the stance chip above.