Auditoria assistida por inteligência artificial com ajustes personalizados e proteção de dados

Arthur Frederico Lerner; Leonardo Flach

doi:10.5585/iptec.v12i2.27075

Authors

Arthur Frederico Lerner Universidade Federal de Santa Catarina https://orcid.org/0000-0002-7027-0562
Leonardo Flach Universidade Federal de Santa Catarina https://orcid.org/0000-0002-4316-0704

DOI:

https://doi.org/10.5585/iptec.v12i2.27075

Keywords:

Auditing, Artificial intelligence, Large language models, Data security, Fine-tuning

Abstract

This project presents an Artificial Intelligence (AI)-assisted auditing solution, utilizing the Large Language Model (LLM) LLama 3, in an offline environment to ensure the security of sensitive data and comply with the General Data Protection Law (LGPD). The developed tool uses personalized prompts to adapt the model to the specific needs of auditors, providing greater flexibility in the process. The tools adopted are open-source, ensuring accessibility and customization for different auditing scenarios. The study explores practical applications, such as data analysis in Excel and PDF, financial indicator calculations, and the identification of accounting anomalies, areas in which the model has proven effective in improving the accuracy and efficiency of the process. Additionally, the offline use offers greater security in handling financial and accounting data, protecting the information from potential leaks. The technical, economic, operational, and legal feasibility was carefully analyzed. The results indicate that the risks associated with implementation are low, with a positive return in terms of efficiency and accuracy in audits. The tool allows for continuous model adaptation through adjustments made directly by auditors, ensuring the solution remains aligned with the specific needs of the context. This project represents a significant advance in the field of auditing, integrating AI in a practical and secure way, with the potential to transform audit execution for companies of various sizes.

Downloads

Author Biographies

Arthur Frederico Lerner, Universidade Federal de Santa Catarina

Researcher and professional in the field of Accounting, with experience in auditing, controllership, and teaching. I have a highly motivated profile for continuous learning and knowledge dissemination, especially in the areas of Finance, Technology, and Data Science. I am currently pursuing a Ph.D. in Accounting at the Federal University of Santa Catarina (UFSC), with a focus on quantitative methods, aiming to further enhance my expertise and contribute to the advancement of the field. I hold a solid academic background, with a master’s degree in Controllership and Accounting from the Federal University of Rio Grande do Sul (UFRGS), completed in 2019, and a bachelor’s degree in Accounting from the same institution, completed in 2016.

Leonardo Flach, Universidade Federal de Santa Catarina

Professor at the Federal University of Santa Catarina, in the Department of Accounting. CNPQ Scientific Productivity Scholar (PQ2). Served as Senior Visiting Professor at the University of Lisbon (Portugal). Holds a postdoctoral degree in Accounting and Finance from the Massachusetts Institute of Technology (MIT/USA). Participates in three Graduate Programs: Graduate Program in Accounting (PPGC), Graduate Program in Intellectual Property and Technology Transfer for Innovation (PROFNIT), and Graduate Program in University Administration.

References

Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., Brynjolfsson, E., Buch, S., Card, D., Castellon, R., Chatterji, N., Chen, A., Creel, K., Davis, J. Q., Demszky, D., Donahue, C., Doumbouya, M., Durmus, E., Ermon, S., Etchemendy, J., Ethayarajh, K., Fei-Fei, L., Finn, C., Gale, T., Gillespie, L., Goel, K., Goodman, N., Grossman, S., Guha, N., Hashimoto, T., Henderson, P., Hewitt, J., Ho, D. E., Hong, J., Hsu, K., Huang, J., Icard, T., Jain, S., Jurafsky, D., Kalluri, P., Karamcheti, S., Keeling, G., Khani, F., Khattab, O., Koh, P. W., Krass, M., Krishna, R., Kuditipudi, R., Kumar, A., Ladhak, F., Lee, M., Lee, T., Leskovec, J., Levent, I., Li, X. L., Li, X., Ma, T., Malik, A., Manning, C. D., Mirchandani, S., Mitchell, E., Munyikwa, Z., Nair, S., Narayan, A., Narayanan, D., Newman, B., Nie, A., Niebles, J. C., Nilforoshan, H., Nyarko, J., Ogut, G., Orr, L., Papadimitriou, I., Park, J. S., Piech, C., Portelance, E., Potts, C., Raghunathan, A., Reich, R., Ren, H., Rong, F., Roohani, Y., Ruiz, C., Ryan, J., Ré, C., Sadigh, D., Sagawa, S., Santhanam, K., Shih, A., Srinivasan, K., Tamkin, A., Taori, R., Thomas, A. W., Tramèr, F., Wang, R. E., Wang, W., Wu, B., Wu, J., Wu, Y., Xie, S. M., Yasunaga, M., You, J., Zaharia, M., Zhang, M., Zhang, T., Zhang, X., Zhang, Y., Zheng, L., Zhou, K., & Liang, P. (2021). On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258. https://doi.org/10.48550/arXiv.2108.07258.

Chen, M., Tworek, J., Jun, H., Yuan, Q., Pinto, H. P. D. O., Kaplan, J., Edwards, H., Burda, Y., Joseph, N., Brockman, G., Ray, A., Puri, R., Krueger, G., Petrov, M., Khlaaf, H., Sastry, G., Mishkin, P., Chan, B., Gray, S., Ryder, N., Pavlov, M., Power, A., Kaiser, L., Bavarian, M., Winter, C., Tillet, P., Petroski Such, F., Cummings, D., Plappert, M., Chantzis, F., Barnes, E., Herbert-Voss, A., Hebgen Guss, W., Nichol, A., Paino, A., Tezak, N., Tang, J., Babuschkin, I., Balaji, S., Jain, S., Saunders, W., Hesse, C., Carr, A. N., Leike, J., Achiam, J., Misra, V., Morikawa, E., Radford, A., Knight, M., Brundage, M., Murati, M., Mayer, K., Welinder, P., McGrew, B., Amodei, D., McCandlish, S., Sutskever, I., & Zaremba, W. (2021). Evaluating large language models trained on code. arXiv preprint arXiv:2107.03374. https://doi.org/10.48550/arXiv.2107.03374.

Dong, M. M., Stratopoulos, T. C., & Wang, V. X. (2024). A scoping review of ChatGPT research in accounting and finance. International Journal of Accounting Information Systems, 55, 100715. https://doi.org/10.1016/j.accinf.2024.100715.

Freitas, M. M., Sallaberry, J. D., Silva, T. B. J., & Rosa, F. S. (2024). Application of Chatgpt 4.0 for solving accounting problems. GCG: revista de globalización, competitividad y gobernabilidad, 18(2), 49-64. DOI: https://doi.org/10.58416/GCG.2024.V18.N2.03

Hillebrand, L., Deußer, T., Dilmaghani, T., Kliem, B., Loitz, R., Bauckhage, C., & Sifa, R. (2022a, August). Kpi-bert: A joint named entity recognition and relation extraction model for financial reports. In 2022 26th International Conference on Pattern Recognition (ICPR) (pp. 606-612). IEEE. https://doi.org/10.1109/ICPR56361.2022.9956191.

Hillebrand, L., Deußer, T., Dilmaghani, T., Kliem, B., Loitz, R., Bauckhage, C., & Sifa, R. (2022b, December). Towards automating numerical consistency checks in financial reports. In 2022 IEEE International Conference on Big Data (Big Data) (pp. 5915-5924). IEEE. https://doi.org/10.1109/BigData55660.2022.10020308.

Kok, T. (2023). Generative LLMs and Textual Analysis in Accounting:(Chat) GPT as Research Assistant? Available at SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4429658.

Longpre, S., Hou, L., Vu, T., Webson, A., Chung, H. W., Tay, Y., Zhou, D., Le, Q. V., Zoph, B., Wei, J., & Roberts, A. (2023). The flan collection: Designing data and methods for effective instruction tuning. arXiv preprint arXiv:2301.13688. Retrieved from https://proceedings.mlr.press/v202/longpre23a.html.

Metaxa, D., Park, J. S., Robertson, R. E., Karahalios, K., Wilson, C., Hancock, J., & Sandvig, C. (2021). Auditing algorithms: Understanding algorithmic systems from the outside in. Foundations and Trends® in Human–Computer Interaction, 14(4), 272-344. http://dx.doi.org/10.1561/1100000083.

Mökander, J., Schuett, J., Kirk, H. R., & Floridi, L. (2023). Auditing large language models: a three-layered approach. AI and Ethics, 1-31. https://doi.org/10.1007/s43681-023-00289-2.

Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., Ray, A., Schulman, J., Hilton, J., Kelton, F., Miller, L., Simens, M., Askell, A., Welinder, P., Christiano, P. F., Leike, J., & Lowe, R. (2022). Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35, 27730-27744. Retrieved from https://proceedings.neurips.cc/paper_files/paper/2022/hash/b1efde53be364a73914f58805a001731-Abstract-Conference.html.

Sarker, I. H. (2024). LLM potentiality and awareness: a position paper from the perspective of trustworthy and responsible AI modeling. Discover Artificial Intelligence, 4(1), 40. https://doi.org/10.1007/s44163-024-00129-0.

Schreyer, M., Sattarov, T., Gierbl, A., Reimer, B., & Borth, D. (2020, October). Learning sampling in financial statement audits using vector quantised variational autoencoder neural networks. Proceedings… First ACM International Conference on AI in Finance (pp. 1-8). https://doi.org/10.1145/3383455.3422546.

Schultz, M., & Tropmann-Frick, M. (2020). Autoencoder neural networks versus external auditors: Detecting unusual journal entries in financial statement audits. Access on 13th November, 2020. Retrieved from https://aisel.aisnet.org/hicss-53/os/risks/3/

Sebastian, G. (2023). Privacy and data protection in ChatGPT and other AI Chatbots: strategies for securing user information. International Journal of Security and Privacy in Pervasive Computing (IJSPPC), 15(1), 1-14. Retrieved from: https://www.igi-global.com/article/privacy-and-data-protection-in-chatgpt-and-other-ai-chatbots/325475

Vasarhelyi, M. A., Moffitt, K. C., Stewart, T., & Sunderland, D. (2023). Large language models: An emerging technology in accounting. Journal of emerging technologies in accounting, 20(2), 1-10. https://doi.org/10.2308/JETA-2023-047.

Wei, J., Wang, X., Schuurmans, D., Bosma, M., Xia, F., Chi, E., Le, Q. V., & Zhou, D. (2022). Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 35, 24824-24837. Retrieved from https://proceedings.neurips.cc/paper_files/paper/2022/hash/9d5609613524ecf4f15af0f7b31abca4-Abstract-Conference.html.