Does Human Collaboration Enhance the Accuracy of Identifying LLM-Generated Deepfake Texts?
Open accessThe DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index.
Abstract
Advances in Large Language Models (e.g., GPT-4, LLaMA) have improved the generation of coherent sentences resembling human writing on a large scale, resulting in the creation of so-called deepfake texts. However, this progress poses security and privacy concerns, necessitating effective solutions for distinguishing deepfake texts from human-written ones. Although prior works studied humans’ ability to detect deepfake texts, none has examined whether “collaboration” among humans improves the detection of deepfake texts. In this study, to address this gap of understanding on deepfake texts, we conducted experiments with two groups: (1) nonexpert individuals from the AMT platform and (2) writing experts from the Upwork platform. The results demonstrate that collaboration among humans can potentially improve the detection of deepfake texts for both groups, increasing detection accuracies by 6.36% for non-experts and 12.76% for experts, respectively, compared to individuals’ detection accuracies. We further analyze the explanations that humans used for detecting a piece of text as deepfake text, and find that the strongest indicator of deepfake texts is their lack of coherence and consistency. Our study provides useful insights for future tools and framework designs to facilitate the collaborative human detection of deepfake texts. The experiment datasets and AMT implementations are available at: https://github.com/huashen218/llm-deepfake-human-study.git
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019) · arXiv · doi:10.48550/arxiv.1910.10683 · Cited by 10,929
Why do humans reason? Arguments for an argumentative theory
Mercier, H., & Sperber, D. (2011) · Behavioral and Brain Sciences · doi:10.1017/s0140525x10000968 · Cited by 2,254
Evaluating Large Language Models Trained on Code
Chen, M., Tworek, J., Jun, H., Yuan, Q., Pinto, H. P. D. O., Kaplan, J., Edwards, H., Burda, Y., Joseph, N., Brockman, G., et al. (2021) · arXiv · doi:10.48550/arxiv.2107.03374 · Cited by 1,436
Defending Against Neural Fake News
Zellers, R., Holtzman, A., Rashkin, H., Bisk, Y., Farhadi, A., Roesner, F., & Choi, Y. (2019) · arXiv · doi:10.48550/arxiv.1905.12616 · Cited by 569 · cited 4×
“However, due to the flaws of stylometric classifiers, deep-learning techniques have been proposed (Bakhtin et al. 2019; Huggingface 2023; Zellers et al. 2019; Ippolito et al. 2020; Ai et al. 2022; Jawa-har, Abdul-Mageed, and Lakshmanan 2022).”
InfluentialEthical and social risks of harm from Language Models
Weidinger, L., Mellor, J., Rauh, M., Griffin, C., Uesato, J., Huang, P.-S., Cheng, M., Glaese, M., Balle, B., Kasirzadeh, A., et al. (2021) · arXiv · doi:10.48550/arxiv.2112.04359 · Cited by 477 · cited 1×
“Moreover, scammers may exploit NTGs to craft sophisticated phishing messages, or stereotyping, misrepresenting, and demeaning content (Weidinger et al. 2021), while malicious code generation (Chen et al. 2021) and disinformation attacks by state-backed operators are also plausible…”
GLTR: Statistical Detection and Visualization of Generated Text
Gehrmann, S., Strobelt, H., & Rush, A. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: System Demonstrations · doi:10.18653/v1/p19-3019 · Cited by 470 · cited 4×
“GLTR color codes words based on the distribution level which improves human performance from 54% to 72% (Gehrmann, Strobelt, and Rush 2019).”
InfluentialGLTR: Statistical Detection and Visualization of Generated Text
Gehrmann, S., Strobelt, H., & Rush, A. M. (2019) · arXiv · doi:10.48550/arxiv.1906.04043 · Cited by 460
All That’s ‘Human’ Is Not Gold: Evaluating Human Evaluation of Generated Text
Clark, E., August, T., Serrano, S., Haduong, N., Gururangan, S., & Smith, N. A. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.565 · Cited by 290 · cited 13×
“The collective body of prior research has consistently high-lighted the inherent di ffi culty involved in solving the deep-fake detection problem (Uchendu et al. 2020; Clark et al. 2021; Ippolito et al. 2020; Dugan et al. 2020; Gehrmann, Strobelt, and Rush 2019).”
Influential
Cited by
A survey on large language model (LLM) security and privacy: The Good, The Bad, and The Ugly
Yao, Y., Duan, J., Xu, K., Cai, Y., Sun, Z., & Zhang, Y. (2024) · High-Confidence Computing · doi:10.1016/j.hcc.2024.100211 · Cited by 1,083
“Countermeasures and detection methods [297, 271, 40, 258, 36, 330, 19, 148, 254] have also been developed to address these emerging issues.”
BackgroundA Survey on LLM-Generated Text Detection: Necessity, Methods, and Future Directions
Wu, J., Yang, S., Zhan, R., Yuan, Y., Wong, D. F., & Chao, L. S. (2023) · arXiv · doi:10.48550/arxiv.2310.14724 · Cited by 212
“Human classification primarily depends on visual observation to discern features indicative of text generation by LLMs. Uchendu et al. [143] noted that a lack of coherence and consistency in LLM-generated text serves as a strong indicator of falsified content.”
BackgroundA Survey on LLM-Generated Text Detection: Necessity, Methods, and Future Directions
Wu, J., Yang, S., Zhan, R., Yuan, Y., Chao, L. S., & Wong, D. F. (2025) · Computational Linguistics · doi:10.1162/coli_a_00549 · Cited by 212
Combating misinformation in the age of LLMs: Opportunities and challenges
Chen, C., & Shu, K. (2024) · AI Magazine · doi:10.1002/aaai.12188 · Cited by 148
Human performance in detecting deepfakes: A systematic review and meta-analysis of 56 papers
Diel, A., Lalgi, T., Schröter, I. C., MacDorman, K. F., Teufel, M., & Bäuerle, A. (2024) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2024.100538 · Cited by 64
“…(Groh et al., 2022), attentional strategies (Bray et al., 2023), feedback training (Holmes et al., 2016), financial incentives (K ¨ obis et al., 2021), human collaboration (Uchendu et al., 2023), creating deepfake carica-tures (Fosco et al., 2022), and raising awareness (Tucciaelli et al., 2022).”
Metadata
- Authors
- Uchendu, A., Lee, J., Shen, H., Le, T., Huang, T.-H. '., & Lee, D.
- Venue
- Proceedings of the AAAI Conference on Human Computation and Crowdsourcing
- Published
- 2023-11-03
- Volume
- 11
- Issue
- 1
- Pages
- 163-174
- ISSN
- 2769-1330 (ISSN-L)
- 2769-1349
- OpenAlex
- W4388333192
- Cited by
- 30
- References
- 26
- FWCI
- 5.67
- Type
- conference paper
- Language
- en
- Open access
- diamond
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Does Human Collaboration Enhance the Accuracy of Identifying LLM-Generated Deepfake Texts? https://app.cochat.ai/papers/10.1609/hcomp.v11i1.27557
- The paper
- Uchendu, A., Lee, J., Shen, H., Le, T., Huang, T.-H. '., & Lee, D. (2023). Does Human Collaboration Enhance the Accuracy of Identifying LLM-Generated Deepfake Texts? Proceedings of the AAAI Conference on Human Computation and Crowdsourcing. https://doi.org/10.1609/hcomp.v11i1.27557
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.