Public Responses to Disaster Warnings from Government Versus AI: Evidence from Behavioral Experiments

Abstract

With the growing integration of artificial intelligence (AI) into warning systems, an important question concerns how the public interprets and responds to warnings delivered by governmental authorities and AI systems. Existing studies have mainly examined single-source effects, while the dynamics of multi-source warnings under conditions of consistency and conflict remain insufficiently explored. Based on two online experiments with residents in China (N = 599), this study investigates the psychological processes of information trust and anticipated regret that shape how the public responds to different warning sources and consistency. The findings show that government-issued warnings generate higher trust than AI warnings, and consistent messages from both sources further enhance trust and protective intentions. In conflicting warning scenarios, anticipated regret becomes prominent, contributing to individuals’ tendency to follow the higher-level warning. The results support a dual-pathway conceptual framework of decision making in multi-source warning contexts, where cognitive trust grounded in institutional authority and technological support coexists with emotional motivation driven by anticipated regret. This study fills an empirical gap in multi-source warning research and offers theoretical and practical insights for building disaster warning systems that integrate institutional credibility with emerging AI technologies.

Keywords

Anticipated regret / Artificial intelligence (AI) / Disaster warning / Government authority / Information trust / Risk communication

References

[1] Allcott H, Gentzkow M. Social media and fake news in the 2016 election. Journal of Economic Perspectives, 2017, 31(2): 211-236

[2] Araujo T, Helberger N, Kruikemeier S, De Vreese CH. In AI we trust? Perceptions about automated decision-making by artificial intelligence. AI & Society, 2020, 35(3): 611-623

[3] Bargain, O., and U. Aminjonov. 2020. Trust and compliance to public health policies in times of COVID-19. Journal of Public Economics 192: Article 104316.

[4] Basher R. Global early warning systems for natural hazards: systematic and people-centred. Philosophical Transactions: Mathematical, Physical and Engineering Sciences, 2006, 364(1845): 2167-2182

[5] Bashkirova, A., and D. Krpan. 2024. Confirmation bias in AI-assisted decision-making: AI triage recommendations congruent with expert judgments increase psychologist trust and recommendation acceptance. Computers in Human Behavior: Artificial Humans 2(1): Article 100066.

[6] Berman, A., K. De Fine Licht, and V. Carlsson. 2024. Trustworthy AI in the public sector: an empirical analysis of a Swedish labor market decision-support system. Technology in Society 76: Article 102471.

[7] Bian Q, Wang L, Xin L, Ma B. Mismatch between warning information and protective behavior: Why experts + AI < 2?. Risk Analysis, 2025, 45(12): 4367-4377

[8] Birch, J. 2021. Science and policy in extremis: The UK’s initial response to COVID-19. European Journal for Philosophy of Science 11(3): Article 90.

[9] Booth K, Harwood A. Insurance as catastrophe: A geography of house and contents insurance in bushfire-prone places. Geoforum, 2016, 69: 44-52

[10] Botzen WJW, Thepaut LD, Banerjee S. Kahneman’s insights for climate risks: lessons from bounded rationality, heuristics and biases. Environmental and Resource Economics, 2025, 88(10): 2663-2688

[11] Cai J, Song C. Do disaster experience and knowledge affect insurance take-up decisions?. Journal of Development Economics, 2017, 124: 83-94

[12] Capraro, V., A. Lentsch, D. Acemoglu, S. Akgun, A. Akhmedova, E. Bilancini, J.-F. Bonnefon, P. Brañas-Garza, et al. 2024. The impact of generative artificial intelligence on socioeconomic inequalities and policy making. PNAS Nexus 3(6): Article 191.

[13] Castelo N, Bos MW, Lehmann DR. Task-dependent algorithm aversion. Journal of Marketing Research, 2019, 56(5): 809-825

[14] Chen H, Greitens SC. Information capacity and social order: The local politics of information integration in China. Governance, 2022, 35(2): 497-523

[15] Clarke L, Chess C, Holmes R, O’Neill KM. Speaking with one voice: risk communication lessons from the US anthrax attacks. Journal of Contingencies and Crisis Management, 2006, 14(3): 160-169

[16] Colombelli S, Carotenuto F, Elia L, Zollo A. Design and implementation of a mobile device app for network-based earthquake early warning systems (EEWSs): application to the PRESTo EEWS in southern Italy. Natural Hazards and Earth System Sciences, 2020, 20(4): 921-931

[17] Dietvorst BJ, Simmons JP, Massey C. Algorithm aversion: people erroneously avoid algorithms after seeing them err. Journal of Experimental Psychology: General, 2015, 144(1): 114-126

[18] Glikson, E., and O. Asscher. 2023. AI-mediated apology in a multilingual work context: implications for perceived authenticity and willingness to forgive. Computers in Human Behavior 140: Article 107592.

[19] Guo, Y., S. An, and T. Comes. 2022. From warning messages to preparedness behavior: the role of risk perception and information interaction in the Covid-19 pandemic. International Journal of Disaster Risk Reduction 73: Article 102871.

[20] Habbal, A., M.K. Ali, and M.A. Abuzaraida. 2024. Artificial intelligence trust, risk and security management (AI TRiSM): frameworks, applications, challenges and future research directions. Expert Systems with Applications 240: Article 122442.

[21] Han Z, Lu X, Hörhager EI, Yan J. The effects of trust in government on earthquake survivors’ risk perception and preparedness in China. Natural Hazards, 2017, 86(1): 437-452

[22] Han, Z., and G. Wu. 2024. Why do people not prepare for disasters? A national survey from China. npj Natural Hazards 1(1): Article 1.

[23] Holroyd TA, Limaye RJ, Gerber JE, Rimal RN, Musci RJ, Brewer J, Sutherland A, Blunt M, et al.. Development of a scale to measure trust in public health authorities: prevalence of trust and association with vaccination. Journal of Health Communication, 2021, 26(4): 272-280

[24] Hou YT-Y, Jung MF. Who is the expert? Reconciling algorithm aversion and algorithm appreciation in AI-supported decision making. Proceedings of the ACM on Human-Computer Interaction, 2021, 5(CSCW2): 1-25

[25] Houston JB, Hawthorne J, Perreault MF, Park EH, Goldstein Hode M, Halliwell MR, Turner McGowen SE, Davis R, et al.. Social media and disasters: a functional framework for social media use in disaster planning, response, and research. Disasters, 2015, 39(1): 1-22

[26] Imran M, Castillo C, Diaz F, Vieweg S. Processing social media messages in mass emergency: a survey. ACM Computing Surveys, 2015, 47(4): 1-38

[27] Jussupow E, Benbasat I, Heinzl A. An integrative perspective on algorithm aversion and appreciation in decision-making. MIS Quarterly, 2024, 48(4): 1575-1590

[28] Kahneman D, Tversky AMacLean LC, Ziemba WT. Prospect theory: An analysis of decision under risk. Handbook of the fundamentals of financial decision making: Part I, 2013, Singapore, World Scientific, 99-127

[29] Kim, D., and J. Kong. 2023. Front-end AI vs. back-end AI: New framework for securing truth in communication during the generative AI era. Frontiers in Communication 8: Article 1243474.

[30] Kim, K.G. 2016. Book review: Deep Learning. Healthcare Informatics Research 22(4): Article 351.

[31] Kulin J, Johansson Sevä I. Who do you trust? How trust in partial and impartial government institutions influences climate policy attitudes. Climate Policy, 2021, 21(1): 33-46

[32] Kuller, M., K. Schoenholzer, and J. Lienert. 2021. Creating effective flood warnings: A framework from a critical review. Journal of Hydrology 602: Article 126708.

[33] LeClerc J, Joslyn S. The cry wolf effect and weather-related decision making. Risk Analysis, 2015, 35(3): 385-395

[34] Lindell MK, Perry RW. The protective action decision model: theoretical modifications and additional evidence. Risk Analysis, 2012, 32(4): 616-632

[35] Liu B, Sundar SS. Should machines express sympathy and empathy? Experiments with a health advice chatbot. Cyberpsychology, Behavior, and Social Networking, 2018, 21(10): 625-636

[36] Liu Z, Wu Y, Feng J. Competition between battery switching and charging in electric vehicle: considering anticipated regret. Environment, Development and Sustainability, 2023, 26(5): 11957-11978

[37] Logg JM, Minson JA, Moore DA. Algorithm appreciation: people prefer algorithmic to human judgment. Organizational Behavior and Human Decision Processes, 2019, 151: 90-103

[38] Lorimer S, McCormack T, Hoerl C, Johnston M, Beck SR, Feeney A. Do both anticipated relief and anticipated regret predict decisions about influenza vaccination?. British Journal of Health Psychology, 2024, 29(1): 134-148

[39] Losee JE, Joslyn S. The need to trust: How features of the forecasted weather influence forecast trust. International Journal of Disaster Risk Reduction, 2018, 30: 95-104

[40] Madan, R., and M. Ashok. 2023. AI adoption and diffusion in public administration: a systematic literature review and future research agenda. Government Information Quarterly 40(1): Article 101774.

[41] Marchezini V, Trajber R, Olivato D, Muñoz VA, De Oliveira Pereira F, Oliveira Luz AE. Participatory early warning systems: youth, citizen science, and intergenerational dialogues on disaster risk reduction in Brazil. International Journal of Disaster Risk Science, 2017, 8(4): 390-401

[42] Morss RE, Mulder KJ, Lazo JK, Demuth JL. How do people perceive, understand, and anticipate responding to flash flood risks and warnings? Results from a public survey in Boulder, Colorado, USA. Journal of Hydrology, 2016, 541: 649-664

[43] Nowak, A., M. Biesaga, K. Ziembowicz, T. Baran, and P. Winkielman. 2023. Subjective consistency increases trust. Scientific Reports 13(1): Article 5657.

[44] Parker JA, Whitmer DE, Sims VK. Warnings for Hurricane Irma: trust of warning type and perceptions of self-efficacy and susceptibility. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2018, 62(1): 1368-1372

[45] Paton, D. 2008. Risk communication and natural hazard mitigation: How trust influences its effectiveness. International Journal of Global Environmental Issues 8(1/2): Article 2.

[46] Peng L, Tan J, Lin L, Xu D. Understanding sustainable disaster mitigation of stakeholder engagement: risk perception, trust in public institutions, and disaster insurance. Sustainable Development, 2019, 27(5): 885-897

[47] Petrova, K., and E.L. Rosvold. 2024. Mitigating the legacy of violence: can flood relief improve people’s trust in government in conflict-affected areas? Evidence from Pakistan. World Development 173: Article 106372.

[48] Reichstein, M., V. Benson, J. Blunk, G. Camps-Valls, F. Creutzig, C.J. Fearnley, B. Han, K. Kornhuber, et al. 2025. Early warning of complex climate risk with integrated artificial intelligence. Nature Communications 16(1): Article 2564.

[49] Robles P, Mallinson DJ. Artificial intelligence technology, public trust, and effective governance. Review of Policy Research, 2025, 42(1): 11-28

[50] Rokhideh M, Fearnley C, Budimir M. Multi-hazard early warning systems in the Sendai framework for disaster risk reduction: achievements, gaps, and future directions. International Journal of Disaster Risk Science, 2025, 16(1): 103-116

[51] Rolnick D, Donti PL, Kaack LH, Kochanski K, Lacoste A, Sankaran K, Ross AS, Milojevic-Dupont N, et al.. Tackling climate change with machine learning. ACM Computing Surveys, 2023, 55(2): 1-96

[52] Shi, P., M. Meier, L. Villiger, K. Tuinstra, P.A. Selvadurai, F. Lanza, S. Yuan, A. Obermann, et al. 2024. From labquakes to megathrusts: Scaling deep learning based pickers over 15 orders of magnitude. Journal of Geophysical Research: Machine Learning and Computation 1(4): Article e2024JH000220.

[53] Shin, D. 2021. The effects of explainability and causability on perception, trust, and acceptance: implications for explainable AI. International Journal of Human-Computer Studies 146: Article 102551.

[54] Sun Y, Hu Q, Grossman S, Basnyat I, Wang P. Comparison of COVID-19 information seeking, trust of information sources, and protective behaviors in China and the US. Journal of Health Communication, 2021, 26(9): 657-666

[55] Tan J, Zhou K, Peng L, Lin L. The role of social networks in relocation induced by climate-related hazards: an empirical investigation in China. Climate and Development, 2022, 14(1): 1-12

[56] Thekdi S, Aven T. Understanding the implications of low knowledge and high uncertainty in risk studies. Risk Analysis, 2024, 44(7): 1651-1665

[57] Thekdi S, Tatar U, Santos J, Chatterjee S. Disaster risk and artificial intelligence: a framework to characterize conceptual synergies and future opportunities. Risk Analysis, 2023, 43(8): 1641-1656

[58] Visave J. Transparency in AI for emergency management: building trust and accountability. AI and Ethics, 2025, 5(4): 3967-3980

[59] Wachinger G, Renn O, Begg C, Kuhlicke C. The risk perception paradox – implications for governance and communication of natural hazards. Risk Analysis, 2013, 33(6): 1049-1065

[60] Weyrich P, Scolobig A, Patt A. Dealing with inconsistent weather warnings: effects on warning quality and intended actions. Meteorological Applications, 2019, 26(4): 569-583

[61] WHO (World Health Organization). Risk communication and community engagement readiness and response to coronavirus disease (COVID-19): Interim guidance, 2020, Geneva, WHO

[62] Wilson, C. 2022. Public engagement and AI: a values analysis of national strategies. Government Information Quarterly 39(1): Article 101652.

[63] Xie, C., H. Gao, Y. Huang, Z. Xue, C. Xu, and K. Dai. 2025. Leveraging the DeepSeek large model: a framework for AI-assisted disaster prevention, mitigation, and emergency response systems. Earthquake Research Advances: Article 100378.

[64] Xu C, Xue Z. Applications and challenges of artificial intelligence in the field of disaster prevention, reduction, and relief. Natural Hazards Research, 2024, 4(1): 169-172

[65] Yang, F., J. Tan, and L. Peng. 2020. The effect of risk perception on the willingness to purchase hazard insurance – A case study in the three gorges reservoir region, China. International Journal of Disaster Risk Reduction 45: Article 101379.

[66] Yoo C, Yoo E, Yan L, Pedraza-Martinez A. Speak with one voice? Examining content coordination and social media engagement during disasters. Information Systems Research, 2024, 35(2): 551-569

[67] Zeelenberg M. Anticipated regret, expected feedback and behavioral decision making. Journal of Behavioral Decision Making, 1999, 12(2): 93-106

[68] Zhan ES, Zheng Q, Dong C, Thorson E. Does AI-generated care-based message increase trust in government? The pivotal role of AI knowledge in government crisis response. International Journal of Strategic Communication, 2025, 19(2): 176-199

[69] Zhang, G., L. Chong, K. Kotovsky, and J. Cagan. 2023. Trust in an AI versus a human teammate: the effects of teammate identity and performance on human-AI cooperation. Computers in Human Behavior 139: Article 107536.

Rights & permissions

The Author(s)

Just Accepted

This article has successfully passed peer review and final editorial review, and will soon enter typesetting, proofreading and other publishing processes. The currently displayed version is the accepted final manuscript. The officially published version will be updated with format, DOI and citation information upon launch. We recommend that you pay attention to subsequent journal notifications and preferentially cite the officially published version. Thank you for your support and cooperation.