{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,3]],"date-time":"2026-11-03T17:34:09Z","timestamp":1793727249693,"version":"4.6.1"},"reference-count":51,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2023,9,25]],"date-time":"2023-09-25T00:00:00Z","timestamp":1695600000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2023,9,25]],"date-time":"2023-09-25T00:00:00Z","timestamp":1695600000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100000780","name":"European Commission","doi-asserted-by":"publisher","award":["101021801"],"award-info":[{"award-number":["101021801"]}],"id":[{"id":"10.13039\/501100000780","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000780","name":"European Commission","doi-asserted-by":"publisher","award":["101070303"],"award-info":[{"award-number":["101070303"]}],"id":[{"id":"10.13039\/501100000780","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title><jats:p>The collection and use of personal data are becoming more common in today\u2019s data-driven culture. While there are many advantages to this, including better decision-making and service delivery, it also poses significant ethical issues around confidentiality and privacy. Text anonymisation tries to prune and\/or mask identifiable information from a text while keeping the remaining content intact to alleviate privacy concerns. Text anonymisation is especially important in industries like healthcare, law, as well as research, where sensitive and personal information is collected, processed, and exchanged under high legal and ethical standards. Although text anonymisation is widely adopted in practice, it continues to face considerable challenges. The most significant challenge is striking a balance between removing information to protect individuals\u2019 privacy while maintaining the text\u2019s usability for future purposes. The question is whether these anonymisation methods sufficiently reduce the risk of re-identification, in which an individual can be identified based on the remaining information in the text. In this work, we challenge the effectiveness of these methods and how we perceive identifiers. We assess the efficacy of these methods against the elephant in the room, the use of AI over big data. While most of the research is focused on identifying and removing personal information, there is limited discussion on whether the remaining information is sufficient to deanonymise individuals and, more precisely, who can do it. To this end, we conduct an experiment using GPT over anonymised texts of famous people to determine whether such trained networks can deanonymise them. The latter allows us to revise these methods and introduce a novel methodology that employs Large Language Models to improve the anonymity of texts.<\/jats:p>","DOI":"10.1038\/s41598-023-42977-3","type":"journal-article","created":{"date-parts":[[2023,9,25]],"date-time":"2023-09-25T12:02:11Z","timestamp":1695643331000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":42,"title":["Man vs the machine in the struggle for effective text anonymisation in the age of large language models"],"prefix":"10.1038","volume":"13","author":[{"given":"Constantinos","family":"Patsakis","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nikolaos","family":"Lykousas","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2023,9,25]]},"reference":[{"key":"42977_CR1","unstructured":"Regulation (EU) 2016\/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95\/46\/EC (General Data Protection Regulation). Official Journal of the European Union, Vol. L119 (4 May 2016) 1\u201388 (2016)."},{"key":"42977_CR2","unstructured":"Samaratiy, P. & Sweeney, L. Protecting privacy when disclosing information: k-anonymity and its enforcement through generalization and suppression (1998)."},{"key":"42977_CR3","unstructured":"Kleinberg, B., Davies, T. & Mozes, M. Textwash-Automated Open-source Text Anonymisation. arXiv preprint arXiv:2208.13081 (2022)."},{"key":"42977_CR4","doi-asserted-by":"publisher","first-page":"24","DOI":"10.1145\/1743546.1743558","volume":"53","author":"A Narayanan","year":"2010","unstructured":"Narayanan, A. & Shmatikov, V. Myths and fallacies of \u201cpersonally identifiable information\u2019\u2019. Commun. ACM 53, 24\u201326 (2010).","journal-title":"Commun. ACM"},{"key":"42977_CR5","unstructured":"Agencia Espa\u00f1ola de Protecci\u00f3n de Datos and European Data Protection Supervisor. 10 misunderstandings related to anonymisation. https:\/\/edps.europa.eu\/data-protection\/our-work\/publications\/papers\/aepd-edps-joint-paper-10-misunderstandings-related_en (2021)."},{"key":"42977_CR6","doi-asserted-by":"publisher","first-page":"51071","DOI":"10.1109\/ACCESS.2020.2980235","volume":"8","author":"A Zigomitros","year":"2020","unstructured":"Zigomitros, A., Casino, F., Solanas, A. & Patsakis, C. A survey on privacy properties for data publishing of relational data. IEEE Access 8, 51071\u201351099 (2020).","journal-title":"IEEE Access"},{"key":"42977_CR7","doi-asserted-by":"crossref","unstructured":"Mamede, N., Baptista, J. & Dias, F. Automated anonymization of text documents. In 2016 IEEE congress on evolutionary computation (CEC) 1287\u20131294 (IEEE, 2016).","DOI":"10.1109\/CEC.2016.7743936"},{"key":"42977_CR8","doi-asserted-by":"crossref","unstructured":"Hassan, F., Domingo-Ferrer, J. & Soria-Comas, J. Anonymization of unstructured data via named-entity recognition. In Modeling Decisions for Artificial Intelligence (eds Torra, V. et al.) 296\u2013305 (Springer International Publishing, Cham, 2018).","DOI":"10.1007\/978-3-030-00202-2_24"},{"key":"42977_CR9","doi-asserted-by":"publisher","first-page":"293","DOI":"10.21105\/joss.00293","volume":"2","author":"B Kleinberg","year":"2017","unstructured":"Kleinberg, B. & Mozes, M. Web-based text anonymization with node.js: Introducing NETANOS (named entity-based text anonymization for open science). J. Open Source Softw. 2, 293. https:\/\/doi.org\/10.21105\/joss.00293 (2017).","journal-title":"J. Open Source Softw."},{"key":"42977_CR10","unstructured":"Sweeney, L. Replacing personally-identifying information in medical records, the scrub system. In Proceedings of the AMIA annual fall symposium 333 (American Medical Informatics Association, 1996)."},{"key":"42977_CR11","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/1472-6947-8-32","volume":"8","author":"I Neamatullah","year":"2008","unstructured":"Neamatullah, I. et al. Automated de-identification of free-text medical records. BMC Med. Inform. Decis. Mak. 8, 1\u201317 (2008).","journal-title":"BMC Med. Inform. Decis. Mak."},{"key":"42977_CR12","unstructured":"Ruch, P., Baud, R.\u00a0H., Rassinoux, A.-M., Bouillon, P. & Robert, G. Medical document anonymization with a semantic lexicon. In Proceedings of the AMIA Symposium 729 (American Medical Informatics Association, 2000)."},{"key":"42977_CR13","doi-asserted-by":"publisher","first-page":"596","DOI":"10.1093\/jamia\/ocw156","volume":"24","author":"F Dernoncourt","year":"2017","unstructured":"Dernoncourt, F., Lee, J. Y., Uzuner, O. & Szolovits, P. De-identification of patient notes with recurrent neural networks. J. Am. Med. Inform. Assoc. 24, 596\u2013606 (2017).","journal-title":"J. Am. Med. Inform. Assoc."},{"key":"42977_CR14","doi-asserted-by":"crossref","unstructured":"Manning, C.\u00a0D. et\u00a0al. The stanford corenlp natural language processing toolkit. In Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations 55\u201360 (2014).","DOI":"10.3115\/v1\/P14-5010"},{"key":"42977_CR15","doi-asserted-by":"publisher","first-page":"S34","DOI":"10.1016\/j.jbi.2017.05.023","volume":"75","author":"Z Liu","year":"2017","unstructured":"Liu, Z., Tang, B., Wang, X. & Chen, Q. De-identification of clinical notes via recurrent neural network and conditional random field. J. Biomed. Inform. 75, S34\u2013S42. https:\/\/doi.org\/10.1016\/j.jbi.2017.05.023 (2017).","journal-title":"J. Biomed. Inform."},{"key":"42977_CR16","doi-asserted-by":"crossref","unstructured":"S\u00e1nchez, D., Batet, M. & Viejo, A. Detecting sensitive information from textual documents: An information-theoretic approach. In Modeling Decisions for Artificial Intelligence: 9th International Conference, MDAI 2012, Girona, Catalonia, Spain, November 21-23, 2012. Proceedings 9 173\u2013184 (Springer, 2012).","DOI":"10.1007\/978-3-642-34620-0_17"},{"key":"42977_CR17","doi-asserted-by":"publisher","first-page":"853","DOI":"10.1109\/TIFS.2013.2239641","volume":"8","author":"D S\u00e1nchez","year":"2013","unstructured":"S\u00e1nchez, D., Batet, M. & Viejo, A. Automatic general-purpose sanitization of textual documents. IEEE Trans. Inf. Forensics Secur. 8, 853\u2013862. https:\/\/doi.org\/10.1109\/TIFS.2013.2239641 (2013).","journal-title":"IEEE Trans. Inf. Forensics Secur."},{"key":"42977_CR18","doi-asserted-by":"publisher","first-page":"189","DOI":"10.1016\/j.jbi.2014.06.008","volume":"52","author":"D S\u00e1nchez","year":"2014","unstructured":"S\u00e1nchez, D., Batet, M. & Viejo, A. Utility-preserving privacy protection of textual healthcare documents. J. Biomed. Inform. 52, 189\u2013198 (2014).","journal-title":"J. Biomed. Inform."},{"key":"42977_CR19","doi-asserted-by":"publisher","unstructured":"\u00d6zlem Uzuner, Sibanda, T.\u00a0C., Luo, Y. & Szolovits, P. A de-identifier for medical discharge summaries. Artif. Intell. Med. 42, 13\u201335. https:\/\/doi.org\/10.1016\/j.artmed.2007.10.001 (2008).","DOI":"10.1016\/j.artmed.2007.10.001"},{"key":"42977_CR20","doi-asserted-by":"publisher","first-page":"342","DOI":"10.1136\/amiajnl-2012-001034","volume":"20","author":"D Carrell","year":"2013","unstructured":"Carrell, D. et al. Hiding in plain sight: Use of realistic surrogates to reduce exposure of protected health information in clinical text. J. Am. Med. Inform. Assoc. 20, 342\u2013348 (2013).","journal-title":"J. Am. Med. Inform. Assoc."},{"key":"42977_CR21","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/1471-2288-10-70","volume":"10","author":"SM Meystre","year":"2010","unstructured":"Meystre, S. M., Friedlin, F. J., South, B. R., Shen, S. & Samore, M. H. Automatic de-identification of textual documents in the electronic health record: A review of recent research. BMC Med. Res. Methodol. 10, 1\u201316 (2010).","journal-title":"BMC Med. Res. Methodol."},{"key":"42977_CR22","doi-asserted-by":"publisher","first-page":"104","DOI":"10.1177\/2515245917749652","volume":"1","author":"CG Walsh","year":"2018","unstructured":"Walsh, C. G. et al. Enabling open-science initiatives in clinical psychology and psychiatry without sacrificing patients\u2019 privacy: Current practices and future challenges. Adv. Methods Pract. Psychol. Sci. 1, 104\u2013114 (2018).","journal-title":"Adv. Methods Pract. Psychol. Sci."},{"key":"42977_CR23","doi-asserted-by":"crossref","unstructured":"Lison, P., Pil\u00e1n, I., S\u00e1nchez, D., Batet, M. & \u00d8vrelid, L. Anonymisation models for text data: State of the art, challenges and future directions. In 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) 4188\u20134203 (2021).","DOI":"10.18653\/v1\/2021.acl-long.323"},{"key":"42977_CR24","doi-asserted-by":"publisher","unstructured":"Devlin, J., Chang, M., Lee, K. & Toutanova, K. BERT: pre-training of deep bidirectional transformers for language understanding. In Burstein, J., Doran, C. & Solorio, T. (eds.) Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT 2019, Minneapolis, MN, USA, June 2-7, 2019, Volume 1 (Long and Short Papers), 4171\u20134186, https:\/\/doi.org\/10.18653\/v1\/n19-1423 (Association for Computational Linguistics, 2019).","DOI":"10.18653\/v1\/n19-1423"},{"key":"42977_CR25","unstructured":"Liu, Y. et\u00a0al. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692 (2019)."},{"key":"42977_CR26","doi-asserted-by":"crossref","unstructured":"Xia, P., Wu, S. & Van\u00a0Durme, B. Which* bert? a survey organizing contextualized encoders. arXiv preprint arXiv:2010.00854 (2020).","DOI":"10.18653\/v1\/2020.emnlp-main.608"},{"key":"42977_CR27","unstructured":"Brants, T., Popat, A.\u00a0C., Xu, P., Och, F.\u00a0J. & Dean, J. Large language models in machine translation. In EMNLP-CoNLL 2007, Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning (ed. Eisner, J.) 858\u2013867 (ACL, 2007)."},{"key":"42977_CR28","unstructured":"Vaswani, A. et\u00a0al. Attention is all you need. In Guyon, I. et\u00a0al. (eds.) Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, December 4\u20139, 2017, Long Beach, CA, USA 5998\u20136008 (2017)."},{"key":"42977_CR29","first-page":"1877","volume":"33","author":"T Brown","year":"2020","unstructured":"Brown, T. et al. Language models are few-shot learners. Adv. Neural Inf. Process. Syst. 33, 1877\u20131901 (2020).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"42977_CR30","first-page":"1315","volume":"100","author":"AD Selbst","year":"2020","unstructured":"Selbst, A. D. Negligence and AI\u2019s human users. BUL Rev. 100, 1315 (2020).","journal-title":"BUL Rev."},{"key":"42977_CR31","doi-asserted-by":"publisher","first-page":"131","DOI":"10.1177\/0968533220945766","volume":"20","author":"H Smith","year":"2020","unstructured":"Smith, H. & Fotheringham, K. Artificial intelligence in clinical decision-making: Rethinking liability. Med. Law Int. 20, 131\u2013154 (2020).","journal-title":"Med. Law Int."},{"key":"42977_CR32","unstructured":"Brundage, M. et\u00a0al. Toward trustworthy AI development: Mechanisms for supporting verifiable claims. arXiv preprint arXiv:2004.07213 (2020)."},{"key":"42977_CR33","unstructured":"European Commission. Proposal for a regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) And Amending Certain Union Legislative Acts. https:\/\/eur-lex.europa.eu\/legal-content\/EN\/TXT\/?uri=celex%3A52021PC0206 (2021)."},{"key":"42977_CR34","unstructured":"GitHub. Copilot. https:\/\/github.com\/features\/copilot (2023)."},{"key":"42977_CR35","unstructured":"OpenAI. Codex. https:\/\/openai.com\/blog\/openai-codex (2023)."},{"key":"42977_CR36","unstructured":"Toulas, B. GitHub Copilot update stops AI model from revealing secrets. https:\/\/www.bleepingcomputer.com\/news\/security\/github-copilot-update-stops-ai-model-from-revealing-secrets\/ (2023)."},{"key":"42977_CR37","unstructured":"OpenAI. API data usage policies. https:\/\/openai.com\/policies\/api-data-usage-policies (2023)."},{"key":"42977_CR38","unstructured":"Ouyang, L. et\u00a0al. Training language models to follow instructions with human feedback. arXiv preprint arXiv:2203.02155 (2022)."},{"key":"42977_CR39","unstructured":"Textwash source code, misclassified sample. https:\/\/github.com\/maximilianmozes\/textwash\/blob\/d347f40cab948bdcb522e3f8829c2a9b05bd7fbc\/paper\/study2\/person_descriptions\/orig\/rowling_69.txt (2023)."},{"key":"42977_CR40","unstructured":"Faker. https:\/\/github.com\/joke2k\/faker."},{"key":"42977_CR41","unstructured":"Osborne, J.\u00a0D. et\u00a0al. Bratsynthetic: Text de-identification using a markov chain replacement strategy for surrogate personal identifying information. arXiv preprint arXiv:2210.16125 (2022)."},{"key":"42977_CR42","unstructured":"Mackey, E., Elliot, M. & O\u2019Hara, K. The Anonymisation Decision-Making Framework (UKAN Publications, 2016)."},{"key":"42977_CR43","unstructured":"spaCy. https:\/\/spacy.io (2023)."},{"key":"42977_CR44","doi-asserted-by":"publisher","unstructured":"Lehman, E., Jain, S., Pichotta, K., Goldberg, Y. & Wallace, B.\u00a0C. Does BERT pretrained on clinical notes reveal sensitive data? In Toutanova, K. et\u00a0al. (eds.) Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT 2021, Online, June 6\u201311, 2021 946\u2013959, https:\/\/doi.org\/10.18653\/v1\/2021.naacl-main.73 (Association for Computational Linguistics, 2021).","DOI":"10.18653\/v1\/2021.naacl-main.73"},{"key":"42977_CR45","doi-asserted-by":"crossref","unstructured":"Brown, H., Lee, K., Mireshghallah, F., Shokri, R. & Tram\u00e8r, F. What does it mean for a language model to preserve privacy? In 2022 ACM Conference on Fairness, Accountability, and Transparency 2280\u20132292 (2022).","DOI":"10.1145\/3531146.3534642"},{"key":"42977_CR46","doi-asserted-by":"publisher","first-page":"313","DOI":"10.1007\/s10676-010-9227-5","volume":"12","author":"M Zimmer","year":"2010","unstructured":"Zimmer, M. \u201cbut the data is already public\u2019\u2019: On the ethics of research in facebook. Ethics Inf. Technol. 12, 313\u2013325 (2010).","journal-title":"Ethics Inf. Technol."},{"key":"42977_CR47","doi-asserted-by":"publisher","unstructured":"Lukas, N. et\u00a0al. Analyzing leakage of personally identifiable information in language models. CoRR. https:\/\/doi.org\/10.48550\/arXiv.2302.00539 (2023). 2302.00539.","DOI":"10.48550\/arXiv.2302.00539"},{"key":"42977_CR48","unstructured":"BLOOM. https:\/\/huggingface.co\/bigscience\/bloom-7b1 (2023)."},{"key":"42977_CR49","unstructured":"Wiggers, K. A year in the making, BigScience\u2019s AI language model is finally available. https:\/\/techcrunch.com\/2022\/07\/12\/a-year-in-the-making-bigsciences-ai-language-model-is-finally-available\/ (2022)."},{"key":"42977_CR50","doi-asserted-by":"publisher","unstructured":"Sur\u00eds, D., Menon, S. & Vondrick, C. Vipergpt: Visual inference via python execution for reasoning. CoRR. https:\/\/doi.org\/10.48550\/arXiv.2303.08128 (2023). 2303.08128.","DOI":"10.48550\/arXiv.2303.08128"},{"key":"42977_CR51","doi-asserted-by":"crossref","unstructured":"Huang, J., Shao, H. & Chang, K.\u00a0C. Are large pre-trained language models leaking your personal information? In Findings of the Association for Computational Linguistics: EMNLP 2022, Abu Dhabi, United Arab Emirates, December 7\u201311, 2022 (eds. Goldberg, Y., et al.) 2038\u20132047 (Association for Computational Linguistics, 2022).","DOI":"10.18653\/v1\/2022.findings-emnlp.148"}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-023-42977-3.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-023-42977-3","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-023-42977-3.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,11,18]],"date-time":"2023-11-18T22:14:57Z","timestamp":1700345697000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-023-42977-3"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,9,25]]},"references-count":51,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2023,12]]}},"alternative-id":["42977"],"URL":"https:\/\/doi.org\/10.1038\/s41598-023-42977-3","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,9,25]]},"assertion":[{"value":"22 March 2023","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"17 September 2023","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"25 September 2023","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"16026"}}