Ethical Challenges in the Development of Virtual Assistants Powered by Large Language Models
Open accessThe DOI matches its Crossref record; 16 of the 16 references it cites were found in a registry or an index.
Abstract
Virtual assistants (VAs) have gained widespread popularity across a wide range of applications, and the integration of Large Language Models (LLMs), such as ChatGPT, has opened up new possibilities for developing even more sophisticated VAs. However, this integration poses new ethical issues and challenges that must be carefully considered, particularly as these systems are increasingly used in public services: transfer of personal data, decision-making transparency, potential biases, and privacy risks. This paper, an extension of the work presented at IberSPEECH 2022, analyzes the current regulatory framework for AI-based VAs in Europe and delves into ethical issues in depth, examining potential benefits and drawbacks of integrating LLMs with VAs. Based on the analysis, this paper argues that the development and use of VAs powered by LLMs should be guided by a set of ethical principles that prioritize transparency, fairness, and harm prevention. The paper presents specific guidelines for the ethical use and development of this technology, including recommendations for data privacy, bias mitigation, and user control. By implementing these guidelines, the potential benefits of VAs powered by LLMs can be fully realized while minimizing the risks of harm and ensuring that ethical considerations are at the forefront of the development process.
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
ChatGPT for Good? On Opportunities and Challenges of Large Language Models for Education
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · doi:10.35542/osf.io/5er8f · Cited by 7,552
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326
“Librispeech: an ASR corpus based on public domain audio books,”
Panayotov, V., Chen, G., Povey, D., & Khudanpur, S. (2015) · 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2015.7178964 · Cited by 6,072
Survey of Hallucination in Natural Language Generation
Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., Ishii, E., Bang, Y. J., Madotto, A., & Fung, P. (2023) · ACM Computing Surveys · doi:10.1145/3571730 · Cited by 5,025 · cited 1×
“Hallucinations are a well-known effect on the responses of Natural Language Generation (NLG) models and LLMs [28].”
BackgroundSurvey of Hallucination in Natural Language Generation
Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., Ishii, E., Bang, Y., Chen, D., Dai, W., et al. (2022) · arXiv · doi:10.48550/arxiv.2202.03629 · Cited by 4,981
Universal Language Model Fine-tuning for Text Classification
Howard, J., & Ruder, S. (2018) · Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/p18-1031 · Cited by 4,100 · cited 1×
“The appearance of Transformers [1] and effective transfer learning methods, such as ULMFiT [2] were the catalysts for models such as Generative Pretrained Transformer (GPT) [3] and Bidirectional Encoder Representations from Transformers (BERT) [4], which, together with unsupervised learning, have…”
Raji, I. D., Smart, A., White, R. N., Mitchell, M., Gebru, T., Hutchinson, B., Smith-Loud, J., Theron, D., & Barnes, P. (2020) · Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency · doi:10.1145/3351095.3372873 · Cited by 2,245 · cited 1×
“…(since there is no collective awareness of the real possibilities or the risks that its use could entail) and, as mentioned in [32], “the industry lacks useful tools and incentives to translate high-level ethics principles to verifiable and actionable criteria for designing and deploying AI” [33].”
BackgroundCommon Voice: A Massively-Multilingual Speech Corpus
Ardila, R., Branson, M., Davis, K., Henretty, M., Kohler, M., Meyer, J., Morais, R., Saunders, L., Tyers, F. M., & Weber, G. (2019) · arXiv · doi:10.48550/arxiv.1912.06670 · Cited by 905 · cited 1×
“For the Speech to Text (STT) module, transfer learning and fine tuning of the XLS-R models [43] is applied using audios from Librispeech [44], from Common voice [45] and audios from an own corpus in Galician and Spanish.”
MethodAnil, R., Dai, A. M., Firat, O., Johnson, M., Lepikhin, D., Passos, A., Shakeri, S., Taropa, E., Bailey, P., Chen, Z., et al. (2023) · arXiv · doi:10.48550/arxiv.2305.10403 · Cited by 834
Cited by
Ghimire, P., Kim, K., & Acharya, M. (2024) · Buildings · doi:10.3390/buildings14010220 · Cited by 175
“Also, the adoption of LLMs raises concerns about creating potential biases [159].”
Saihi, A., Ben-Daya, M., Hariga, M., & As'ad, R. (2024) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2024.100274 · Cited by 75
Future Shock: Generative AI and the International AI Policy and Governance Crisis
Leslie, D., & Perini, A. M. (2024) · Harvard Data Science Review · doi:10.1162/99608f92.88b4cc98 · Cited by 33
Context Is King: Large Language Models’ Interpretability in Divergent Knowledge Scenarios
Piñeiro-Martín, A., Santos-Criado, F.-J., García-Mateo, C., Docío-Fernández, L., & López-Pérez, M. D. C. (2025) · Applied Sciences · doi:10.3390/app15031192 · Cited by 4
Self-citation
Metadata
- Authors
- Piñeiro-Martín, A., García-Mateo, C., Docío-Fernández, L., & López-Pérez, M. D. C.
- Venue
- Electronics
- Publisher
- MDPI AG
- Published
- 2023-07-21
- Volume
- 12
- Issue
- 14
- Pages
- 3170
- ISSN
- 2079-9292 (ISSN-L)
- OpenAlex
- W4385216086
- Cited by
- 64
- References
- 16
- FWCI
- 15.63
- Type
- journal article
- Language
- en
- Open access
- gold
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
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.