This study proposes a new framework based on causal inference for cross-domain risk assessment, with specific focus on the problem of transferring predictive models from financial credit to health insurance. The proposed methodology employs directed acyclic graphs to define domain-invariant causal relationships and federated learning architectures to preserve data privacy. With 85,476 financial credit records and 62,318 health insurance records, the framework combines causal discovery algorithms with ensemble learning approaches to build robust risk assessment models. When applied in financial credit assessment, the causal model reaches an area under the curve (AUC) of 0.892 and an F1-score of 0.724 and retains a performance retention rate of 95.4% when transferred to the health insurance sector (AUC = 0.851), and markedly outperforms legacy machine learning methods that face an average performance loss of 15.2%. Causal feature consistency is at 0.91, describing consistent risk relationships under varying domains. The model shows only 3.2% performance variation under changing economic conditions, compared to an 11.7% variation realised with conventional methods. This research explains that causal inference lays a sound methodological foundation for building transferable risk assessment models, presenting measurable gains for organisations that seek to leverage insights across industries while maintaining compliance with regulatory requirements and data privacy legislations.

Luqing Ren. Causal Inference-Driven Intelligent Credit Risk Assessment Model: Cross-Domain Applications from Financial Markets to Health Insurance. Academic Journal of Computing & Information Science (2025), Vol. 8, Issue 8: 8-14. https://doi.org/10.25236/AJCIS.2025.080802.

[1] Zhang, X., & Yu, L. (2024). Consumer credit risk assessment: A review from the state - of - the - art classification algorithms, data traits, and learning methods. Expert Systems with Applications, 237, 121484.

[2] Magliacane, S., Van Ommen, T., Claassen, T., Bongers, S., Versteeg, P., & Mooij, J. M. (2018). Domain adaptation by using causal inference to predict invariant conditional distributions. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS, 31.

[3] Zhang, K., Gong, M., Stojanov, P., Huang, B., Liu, Q., & Glymour, C. (2020). Domain adaptation as a problem of inference on graphical models. In ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS (NeurIPS 2020).

[4] Li, Y., Chen, X., & Wang, H. (2024). Privacy - enhancing collaborative information sharing through federated learning – A case of the insurance industry. arXiv preprint arXiv:2402.14983.

[5] Kumar, A., & Singh, P. (2024). Predicting cross - selling health insurance products using machine - learning techniques. JOURNAL OF COMPUTER INFORMATION SYSTEMS, 64(8), 1045 - 1062.

[6] Stojanov, P., Gong, M., Carbonell, J., & Zhang, K. (2022). Credit risk modeling using transfer learning and domain adaptation. FRONTIERS IN BIG DATA, 5, 852649.

[7] Floch, L., & Castellani, M. (2022). Measuring the model risk - adjusted performance of machine learning algorithms in credit default prediction. FINANCIAL INNOVATION, 8(1), 1 - 35.

[8] Emmanuel, I., Sun, Y., & Wang, Z. (2024). A machine learning - based credit risk prediction engine system using a stacked classifier and a filter - based feature selection method. JOURNAL OF BIG DATA, 11, 23.

[9] Rieke, N., Hancox, J., Li, W., Milletari, F., Roth, H. R., Albarqouni, S., Bakas, S., Galtier, M. N., Landman, B. A., Maier - Hein, K., Ourselin, S., Sheller, M., Summers, R. M., Trask, A., Xu, D., Baust, M., & Cardoso, M. J. (2020). The future of digital health with federated learning. npj DIGITAL MEDICINE, 3(1), 119.

[10] Quan, Z., Wang, Z., Gan, G., & Valdez, E. A. (2024). Automated machine learning in insurance. arXiv preprint arXiv:2408.14331.