Governing Artificial Intelligence in Catastrophe Risk Intelligence: The TRUST Framework for Trustworthy and Examination-Ready AI Governance in U.S. Property and Casualty Insurance

Abstract

Artificial intelligence (AI) is becoming an operating layer beneath underwriting, pricing, catastrophe risk assessment, and claims administration in the U.S. Property and Casualty (P&C) insurance sector, yet the governance frameworks available to insurers remain either sector-agnostic or limited to compliance-level guidance. This paper introduces the Transparency, Risk ownership, Underlying data integrity, Surveillance, and Traceability (TRUST) Framework, a five-pillar, system-level governance model developed specifically for AI deployed in catastrophe-exposed insurance operations, together with three complementary original contributions: an AI Governance Maturity Model spanning five organizational stages, a Catastrophe AI Governance Lifecycle describing the continuous operational cycle through which TRUST is applied, and an Executive Decision Model linking governance quality to enterprise and national resilience outcomes. Drawing on literature spanning AI governance, explainable AI, algorithmic accountability, model risk management, catastrophe risk modelling, insurance regulation, operational resilience, and model lifecycle engineering, the paper identifies a research gap: existing frameworks address governance principles or regulatory compliance in the abstract but do not integrate system-level AI governance with the specific non-stationary risk conditions and regulatory examination trajectory characteristic of catastrophe-exposed insurance. Using qualitative conceptual framework development, comparative regulatory analysis, and case-study illustration grounded in the market disruption following Hurricane Ian, the paper positions TRUST as a framework that operationalizes, rather than replaces, existing standards including the NIST AI Risk Management Framework, ISO/IEC 42001, the OECD AI Principles, the NAIC Model Bulletin, COSO Enterprise Risk Management, and Federal Reserve model risk guidance. Limitations, practical implications, and directions for future empirical research are discussed.

REFERENCES

[1] B. D. Mittelstadt, P. Allo, M. Taddeo, S. Wachter, and L. Floridi, “The ethics of algorithms: Mapping the debate,” Big Data & Society, vol. 3, no. 2, 2016.

[2] M. Ananny and K. Crawford, “Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability,” New Media & Society, vol. 20, no. 3, pp. 973–989, 2018.

[3] I. D. Raji, A. Smart, R. N. White, M. Mitchell, T. Gebru, B. Hutchinson, J. Smith-Loud, D. Theron, and P. Barnes, “Closing the AI accountability gap: Defining an end-to-end framework for internal algorithmic auditing,” in Proc. 2020 Conf. Fairness, Accountability, and Transparency (FAT* '20), Barcelona, Spain, 2020, pp. 33–44.

[4] S. M. Lundberg and S.-I. Lee, “A unified approach to interpreting model predictions,” in Advances in Neural Information Processing Systems (NeurIPS), vol. 30, 2017.

[5] M. T. Ribeiro, S. Singh, and C. Guestrin, “‘Why should I trust you?’: Explaining the predictions of any classifier,” in Proc. 22nd ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining (KDD), 2016, pp. 1135–1144.

[6] F. Doshi-Velez and B. Kim, “Towards a rigorous science of interpretable machine learning,” arXiv preprint arXiv:1702.08608, 2017.

[7] S. Wachter, B. Mittelstadt, and L. Floridi, “Why a right to explanation of automated decision-making does not exist in the General Data Protection Regulation,” International Data Privacy Law, vol. 7, no. 2, pp. 76–99, 2017.

[8] L. Floridi et al., “AI4People—An ethical framework for a good AI society: Opportunities, risks, principles, and recommendations,” Minds and Machines, vol. 28, no. 4, pp. 689–707, 2018.

[9] A. Jobin, M. Ienca, and E. Vayena, “The global landscape of AI ethics guidelines,” Nature Machine Intelligence, vol. 1, pp. 389–399, 2019.

[10] S. Barocas and A. D. Selbst, “Big data’s disparate impact,” California Law Review, vol. 104, pp. 671–732, 2016.

[11] A. D. Selbst and S. Barocas, “The intuitive appeal of explainable machines,” Fordham Law Review, vol. 87, pp. 1085–1139, 2018.

[12] J. A. Kroll, J. Huey, S. Barocas, E. W. Felten, J. R. Reidenberg, D. G. Robinson, and H. Yu, “Accountable algorithms,” University of Pennsylvania Law Review, vol. 165, no. 3, pp. 633–705, 2017.

[13] C. Rudin, “Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead,” Nature Machine Intelligence, vol. 1, pp. 206–215, 2019.

[14] N. Diakopoulos, “Accountability in algorithmic decision making,” Communications of the ACM, vol. 59, no. 2, pp. 56–62, 2016.

[15] M. Veale and R. Binns, “Fairer machine learning in the real world: Mitigating discrimination without collecting sensitive data,” Big Data & Society, vol. 4, no. 2, 2017.

[16] H. Kunreuther and E. Michel-Kerjan, At War with the Weather: Managing Large-Scale Risks in a New Era of Catastrophes. Cambridge, MA, USA: MIT Press, 2009.

[17] P. Grossi and H. Kunreuther, Eds., Catastrophe Modeling: A New Approach to Managing Risk. New York, NY, USA: Springer, 2005.

[18] National Institute of Standards and Technology, Artificial Intelligence Risk Management Framework (AI RMF 1.0), U.S. Dept. of Commerce, Gaithersburg, MD, USA, 2023.

[19] International Organization for Standardization, ISO/IEC 42001:2023—Information Technology—Artificial Intelligence—Management System, ISO, Geneva, Switzerland, 2023.

[20] Organisation for Economic Co-operation and Development, OECD Principles on Artificial Intelligence, OECD, Paris, France, 2019.

[21] National Association of Insurance Commissioners, Model Bulletin: Use of Artificial Intelligence Systems by Insurers, NAIC, Kansas City, MO, USA, Dec. 2023.

[22] Kennedys Law LLP, “Understanding the NAIC model AI bulletin: What it means for insurers,” Jan. 21, 2025.

[23] Fenwick, “NAIC expands AI Systems Evaluation Tool pilot program to 12 states: Key updates for insurers and AI vendors supporting insurers,” Mar. 9, 2026.

[24] Crowell & Moring LLP, “NAIC intensifies AI regulatory focus: What health insurance payors need to know,” Mar. 25, 2026.

[25] Board of Governors of the Federal Reserve System and Office of the Comptroller of the Currency, Supervisory Guidance on Model Risk Management (SR 11-7), Washington, DC, USA, 2011.

[26] Committee of Sponsoring Organizations of the Treadway Commission (COSO), Enterprise Risk Management—Integrating with Strategy and Performance, 2017.

[27] Swiss Re Institute, “sigma 1/2024: Natural catastrophes in 2023—gearing up for today’s and tomorrow’s weather risks,” Zurich, Switzerland, Mar. 2024.

[28] Swiss Re Institute, “sigma 1/2025: Natural catastrophes: insured losses on trend to USD 145 billion in 2025,” Zurich, Switzerland, Apr. 2025.

[29] Swiss Re Institute, “sigma 1/2026: Natural catastrophes in 2025—the persistent rise of wildfire and storm risk,” Zurich, Switzerland, 2026.

[30] National Centers for Environmental Information, “Billion-dollar weather and climate disasters,” National Oceanic and Atmospheric Administration (NOAA), Asheville, NC, USA, 2025.

[31] McKinsey & Company, “The state of AI: Global survey,” QuantumBlack, AI by McKinsey, Nov. 2025.

[32] B. Mittelstadt, “Principles alone cannot guarantee ethical AI,” Nature Machine Intelligence, vol. 1, no. 11, pp. 501–507, 2019.

[33] European Parliament and Council of the European Union, Regulation (EU) 2024/1689 Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act), Official Journal of the European Union, Jul. 2024.

[34] The Institute of Internal Auditors, The IIA’s Three Lines Model: An Update of the Three Lines of Defense, IIA, Lake Mary, FL, USA, 2020.

[35] M. Eulerich and A. Eulerich, “The new three lines model for structuring corporate governance—A critical discussion of similarities and differences,” Corporate Ownership & Control, vol. 18, no. 2, pp. 218–225, 2020.

[36] G. Woo, Calculating Catastrophe. London, U.K.: Imperial College Press, 2011.

[37] B. Keller, Promoting Responsible Artificial Intelligence in Insurance, The Geneva Association, Zurich, Switzerland, 2020.

[38] I. Flückiger and K.-U. Schanz, Regulation of Artificial Intelligence in Insurance: Balancing Consumer Protection and Innovation, The Geneva Association, Zurich, Switzerland, 2023.

[39] Deloitte, “The AI imperative in insurance,” presented to the Federal Advisory Committee on Insurance, U.S. Dept. of the Treasury, Washington, DC, USA, Mar. 20, 2024.

[40] Basel Committee on Banking Supervision, Principles for Operational Resilience, Bank for International Settlements, Basel, Switzerland, Mar. 2021.

[41] B. Green and Y. Chen, “The principles and limits of algorithm-in-the-loop decision making,” Proceedings of the ACM on Human-Computer Interaction, vol. 3, no. CSCW, Article 50, pp. 1–24, 2019.

[42] S. Amershi, A. Begel, C. Bird, R. DeLine, H. Gall, E. Kamar, N. Nagappan, B. Nushi, and T. Zimmermann, “Software engineering for machine learning: A case study,” in Proc. 2019 IEEE/ACM 41st Int. Conf. Software Engineering: Software Engineering in Practice (ICSE-SEIP), Montreal, Canada, 2019, pp. 291–300.

[43] Swiss Re Institute, “Hurricane Ian drives natural catastrophe year-to-date insured losses to USD 115 billion,” Zurich, Switzerland, Dec. 1, 2022.

[44] State of Florida Office of Insurance Regulation, “Hurricane Ian catastrophe claims data,” floir.gov, 2024.

[45] Older & Lundy, “Hurricane Ian insurance claim deadline,” May 23, 2024.

[46] United Policyholders, “Resolving Hurricane Ian insurance disputes,” Oct. 16, 2023.