The DOI matches its Crossref record; 69 of the 73 references it cites were found in a registry or an index.
Bad Characters: Imperceptible NLP Attacks
Open accessAbstract
Several years of research have shown that machine-learning systems are vulnerable to adversarial examples, both in theory and in practice. Until now, such attacks have primarily targeted visual models, exploiting the gap between human and machine perception. Although text-based models have also been attacked with adversarial examples, such attacks struggled to preserve semantic meaning and indistinguishability. In this paper, we explore a large class of adversarial examples that can be used to attack text-based models in a black-box setting without making any human-perceptible visual modification to inputs. We use encoding-specific perturbations that are imperceptible to the human eye to manipulate the outputs of a wide range of Natural Language Processing (NLP) systems from neural machine-translation pipelines to web search engines. We find that with a single imperceptible encoding injection – representing one invisible character, homoglyph, reordering, or deletion – an attacker can significantly reduce the performance of vulnerable models, and with three injections most models can be functionally broken. Our attacks work against currently-deployed commercial systems, including those produced by Microsoft and Google, in addition to open source models published by Facebook, IBM, and HuggingFace. This novel series of attacks presents a significant threat to many language processing systems: an attacker can affect systems in a targeted manner without any assumptions about the underlying model. We conclude that text-based NLP systems require careful input sanitization, just like conventional applications, and that given such systems are now being deployed rapidly at scale, the urgent attention of architects and operators is required.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The text we found for this paper looks like another document, so the self-check wasn’t run.
Citations
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148
Exploiting Generative AI to Scale up Intelligent Tutoring Systems
Jakubův, J., Chvalovský, K., Goertzel, Z., Kaliszyk, C., Olšák, M., Piotrowski, B., Schulz, S., Suda, M., & Urban, J. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.itp.2023.19 · Cited by 76,658
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
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325
Storn, R., & Price, K. (1997) · Journal of Global Optimization · doi:10.1023/a:1008202821328 · Cited by 29,569
Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation
Cho, K., van Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014) · Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1179 · Cited by 25,545
Sequence to Sequence Learning with Neural Networks
Sutskever, I., Vinyals, O., & Le, Q. V. (2014) · arXiv · doi:10.48550/arxiv.1409.3215 · Cited by 17,766
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V. (2019) · arXiv · doi:10.48550/arxiv.1907.11692 · Cited by 16,515
Explaining and Harnessing Adversarial Examples
Goodfellow, I. J., Shlens, J., & Szegedy, C. (2014) · arXiv · doi:10.48550/arxiv.1412.6572 · Cited by 10,381
Research works printed but not linked: 10
- 2020, Unicode Bidirectional Algorithm, The Unicode Consortium
- 2014, Unicode Security Considerations, The Unicode Consortium
- 2015, CAPEC-632: Homograph Attack via Homoglyphs (Version 3.4), MITRE Common Attack Pattern Enumeration and Classification 632
- knight, 2018, Evasion with unicode format characters, SpamAssassin - Dev
- 10.18653/v1/D17-1026 · DOI
- savani, 2021, DistilBERT Base Uncased Emotion
- weaver, 1949, Translation, Translation Machine Translation of Languages Fourteen Essays
- 2020, Münchener Digitalisierungszentrum (MDZ) - Bayerische Staatsbibliothek, BERT Large Cased Finetuned CoNLL03 English
- frenkel, 2020, Facebook is failing in global disinformation fight, says former worker, New York Times
- 2020, Unicode Security Considerations, The Unicode Consortium
Cited by
Shen, X., Chen, Z., Backes, M., Shen, Y., & Zhang, Y. (2023) · arXiv · doi:10.48550/arxiv.2308.03825 · Cited by 861
Greshake, K., Abdelnabi, S., Mishra, S., Endres, C., Holz, T., & Fritz, M. (2023) · Proceedings of the 16th ACM Workshop on Artificial Intelligence and Security · doi:10.1145/3605764.3623985 · Cited by 667
Shen, X., Chen, Z., Backes, M., Shen, Y., & Zhang, Y. (2024) · Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3658644.3670388 · Cited by 441
Adversarial machine learning : a taxonomy and terminology of attacks and mitigations
Vassilev, A., Oprea, A., Fordyce, A. J., & Anderson, H. S. (2024) · doi:10.6028/nist.ai.100-2e2023 · Cited by 317
A Watermark for Large Language Models
Kirchenbauer, J., Geiping, J., Wen, Y., Katz, J., Miers, I., & Goldstein, T. (2023) · arXiv · doi:10.48550/arxiv.2301.10226 · Cited by 194
“Like homoglyphs, these characters must be removed through canonicalization (Pajola & Conti, 2021; Boucher et al., 2022).”
Metadata
- Authors
- Boucher, N., Shumailov, I., Anderson, R., & Papernot, N.
- Affiliations
- University of Cambridge (GB)
- Venue
- 2022 IEEE Symposium on Security and Privacy (SP)
- Publisher
- IEEE
- Published
- 2022-05
- Pages
- 1987-2004
- ISSN
- 1081-6011 (ISSN-L)
- OpenAlex
- W3176393001
- Cited by
- 102
- References
- 73
- Citations by year
- 2026: 13 · 2025: 26 · 2024: 23 · 2023: 32 · 2022: 5
- FWCI
- 11.89
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE Symposium on Security and Privacy (S&P) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Edinburgh Research Explorer (University of Edinburgh) (submittedVersion, cc-by)
- arXiv (Cornell University)
- Funders
- Canadian Institute for Advanced Research
- Bosch-Forschungsstiftung
- Defense Advanced Research Projects Agency
- Natural Sciences and Engineering Research Council of Canada
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
- Physical Sciences › Computer Science › Hardware and Architecture › Security and Verification in Computing
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Keywords
- adversarial examples, imperceptible attacks, black-box attacks, homoglyphs, neural machine translation, search engines, input sanitization, semantic preservation, large language models
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: Bad Characters: Imperceptible NLP Attacks. https://app.cochat.ai/papers/10.1109/sp46214.2022.9833641
- The paper
- Boucher, N., Shumailov, I., Anderson, R., & Papernot, N. (2022). Bad Characters: Imperceptible NLP Attacks. 2022 IEEE Symposium on Security and Privacy (SP). https://doi.org/10.1109/sp46214.2022.9833641
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.