Architecting Trustworthy LLMs: A Unified TRUST Framework for Mitigating AI Hallucination
DOI:
https://doi.org/10.63313/JCSFT.9019Keywords:
AI hallucination, large language models, trustworthy AI, TRUST model, retrieval-augmented generation (RAG), human-AI collaborationAbstract
The use of large language models (LLMs) in high-stakes domains like engineer-ing and public policy is hindered by their tendency to hallucinate—generating factually incorrect or unverifiable information. Although technical mitigations such as Retrieval-Augmented Generation (RAG) and Reinforcement Learning from Human Feedback (RLHF) exist, they are typically applied as isolated point-solutions, lacking a unifying framework. This paper analyses the archi-tectural roots of AI hallucination, classifying them into cognitive gaps and sta-tistical biases, and then shows the shortcomings in current mitigation strategies. To address this, we introduce the TRUST model, which is a systematic structure built on five interconnected pillars: Transparency, Reliability, Uncertainty, Su-pervision, and Traceability. The TRUST framework serves as the architectural basis to orchestrate existing technologies and establish accountability in the lifecycle of AI and to build robust and accountable AIs for vital societal applications [2].
References
[1] Arora, S., & Narayan, P. (2023). A survey on hallucination in large language models. arXiv preprint arXiv:2311.05232. https://arxiv.org/abs/2311.05232
[2] Marco Ramponi. (2023). The full story of large language models and RLHF. https://[2].com/blog/the-full-story-of-large-language-models-and-rlhf
[3] Bai, Y., Zhan, H., & Zhang, W. (2023). A survey of reinforcement learning from human feedback. arXiv preprint arXiv:2312.14925. https://arxiv.org/abs/2312.14925
[4] Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of sto-chastic parrots: Can language models be too big? In Proceedings of the 2021 ACM Confer-ence on Fairness, Accountability, and Transparency (FAccT ’21) (pp. 610–623). ACM.
[5] Chen, L., Zhang, Y., & Wang, M. (2024). AI hallucination: Towards a comprehensive classi-fication of distorted information within AIGC. Humanities and Social Sciences Communi-cations, 11(1), 345. https://www.nature.com/articles/s41599-024-03811-x
[6] Floridi, L., & Chiriatti, M. (2020). GPT-3: Its nature, scope, limits, and consequences. Minds and Machines, 30(4), 681–694. https://doi.org/10.1007/s11023-020-09548-1
[7] Gao, T., Zhang, R., & Zhang, Y. (2023). Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997. https://arxiv.org/abs/2312.10997
[8] Ivan Belcic. (2024). What is retrieval augmented generation (RAG)? https://www.ibm.com/think/topics/retrieval-augmented-generation
[9] Lambert N, Werra L. (2022). Illustrating reinforcement learning from human feedback (RLHF). https://huggingface.co/blog/rlhf
[10] Joshi, A., & Huang, J. (2024). AI hallucinations: A misnomer worth clarifying. arXiv preprint arXiv:2401.06796. https://arxiv.org/pdf/2401.06796
[11] Ji, J., et al. (2023). Survey of Hallucination in Natural Language Generation. ACM Comput. Surv., *55*, 12, Article 248. https://doi.org/10.1145/3571730
[12] Kim, S. H., & Park, E. (2024). Detecting hallucinations in large language models using se-mantic uncertainty estimators. Nature, 630, 112–118. https://www.nature.com/articles/s41586-024-07421-0
[13] Kuhn, L., Gal, Y., & Farquhar, S. (2023). Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generation. arXiv preprint arXiv:2302.09664.
[14] Lewis, P., Perez, E., Piktus, A., et al. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. arXiv preprint arXiv:2005.11401. https://arxiv.org/abs/2005.11401
[15] Li, H., & Xu, Q. (2023). Secrets of RLHF in large language models: PPO and reward model-ing. arXiv preprint arXiv:2307.04964. https://arxiv.org/abs/2307.04964
[16] Li, M., & Wang, Z. (2024). Benchmarking large language models in retrieval-augmented generation. Proceedings of the AAAI Conference on Artificial Intelligence, 38(2), 2234–2243. https://ojs.aaai.org/index.php/AAAI/article/view/29728
[17] Lightman, H., et al. (2024). Let's Verify Step by Step. arXiv preprint arXiv:2305.20050.
[18] Lin, Z., Trivedi, S., & Sun, J. (2024). Generating with Confidence: Uncertainty Quantification for Black-Box Large Language Models. arXiv preprint arXiv:2405.16728.
[19] Li, Y., et al. (2024). HaluBench: A Large-Scale Hallucination Evaluation Benchmark for Large Language Models. arXiv preprint arXiv:2401.06341.
[20] Michał Oleszak. (2024). Reinforcement learning from human feedback for large language models. https://neptune.ai/blog/reinforcement-learning-from-human-feedback-for-llms
[21] Rick Merritt. (2025). What is retrieval-augmented generation (RAG)? https://blogs.nvidia.com/blog/what-is-retrieval-augmented-generation
[22] Ouyang, L., Wu, J., Jiang, X., et al. (2022). Training language models to follow instructions with human feedback. arXiv preprint arXiv:2203.02155. https://arxiv.org/abs/2203.02155
[23] Rahman, F., & Singh, A. (2024). A comprehensive survey of hallucination mitigation tech-niques in large language models. arXiv preprint arXiv:2401.01313. https://arxiv.org/abs/2401.01313
[24] Raji, I. D., Bender, E. M., Paullada, A., Denton, E., & Hanna, A. (2021). AI and the every-thing-in-the-whole-wide-world benchmark. In Proceedings of the NeurIPS Datasets and Benchmarks Track (Round 1).
[25] Wang, Ziwei and Zhong, Jiachen and Zhu, Di, (2025). A Comprehensive Review of Large AI Model Applications in Lung Cancer, From Screening to Treatment Planning. http://dx.doi.org/10.2139/ssrn.5633170
[26] Wu, C., & Lin, Z. (2023). Artificial intelligence hallucinations: A misrepresentation of factu-ality in AI-generated content. Frontiers in Artificial Intelligence, 6, 1196. https://pmc.ncbi.nlm.nih.gov/articles/PMC9939079/
[27] Yang, Q., Liu, Y., Chen, T., & Tong, Y. (2019). Federated Machine Learning: Concept and Ap-plications. ACM Transactions on Intelligent Systems and Technology (TIST), 10(2), 1-19.
[28] Yao, J., & Han, X. (2024). The problem of AI hallucination and how to solve it. Europea Conference on e-Learning (ECEL). https://papers.academic-conferences.org/index.php/ecel/article/view/2584
Downloads
Published
Issue
Section
License
Copyright (c) 2025 by author(s) and Erytis Publishing Limited.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
How to Cite
