AI焦虑的安全管理困境:人机信任与系统透明度机制*

doi:10.16265/j.cnki.issn1003-3033.2026.03.0299

摘要/Abstract

摘要：

为解决人工智能(AI)在企业安全管理与决策实践中因系统复杂性提升、信息不确定性增强而诱发的员工AI焦虑问题,本文基于不确定管理理论(UMT),构建“AI焦虑-人机信任-人机协同决策质量”的机制模型,并引入系统透明度作为边界条件,解释在风险提示与算法黑箱等情境下员工对AI的威胁评估与应对方式;同时以523名AI应用企业员工为样本开展问卷调查,采用验证性因子分析(CFA)与结构方程模型对测量模型、路径关系及调节效应进行检验,并控制性别、年龄、教育、岗位性质与AI使用频率等变量。结果显示:AI焦虑降低人机协同决策质量;人机信任在AI焦虑与协同决策质量之间发挥部分中介作用;系统透明度正向调节人机信任对协同决策质量的影响,高透明度可促进信任向高质量协同转化。

关键词: 人工智能(AI)焦虑, 安全管理, 人机信任, 系统透明度, 人机协同决策质量, 不确定管理理论(UMT)

Abstract:

To address employee' AI anxiety arising from increased system complexity and heightened information uncertainty in the application of AI to corporate safety management and decision-making, a mechanism model of "AI anxiety-human-AI trust-human-AI collaborative decision quality" was developed based on UMT. It introduced system transparency as a boundary condition to explain how employees appraised and coped with AI-related threats in contexts such as risk warnings and algorithmic black boxes. A questionnaire survey was conducted with a sample of 523 employees from AI-adopting enterprises. Confirmatory factor analysis(CFA) and structural equation modeling were employed to test the measurement model, path relationships, and moderating effects, while controlling for variables such as gender, age, education, job type, and AI usage frequency. The results show that AI anxiety reduces the quality of human-AI collaborative decision-making. Human-AI trust partially mediates the relationship between AI anxiety and collaborative decision-making quality. System transparency positively moderates the effect of human-AI trust on collaborative decision-making quality, such that higher transparency facilitates the translation of trust into higher-quality collaboration.

Key words: artificial intelligence (AI) anxiety, safety management, human-AI trust, system transparency, human-AI collaborative decision quality, uncertainty management theory(UMT)

中图分类号:

X912.9安全人机学

牛莉霞, 李波, 李国. AI焦虑的安全管理困境:人机信任与系统透明度机制^*[J]. 中国安全科学学报, 2026, 36(3): 33-40.

NIU Lixia, LI Bo, LI Guo. Safety management dilemma under AI anxiety: human-AI trust and system transparency mechanism[J]. China Safety Science Journal, 2026, 36(3): 33-40.

图/表 8

图1

表1

表2

表3

表4

表5

图2

表6

参考文献 31

[1]	王大龙, 王冰山, 曹睿, 等. 煤矿安全管理行为交互机制与模型[J]. 中国安全科学学报, 2025, 35(5): 32-38. doi: 10.16265/j.cnki.issn1003-3033.2025.05.1094
	WANG Dalong, WANG Bingshan, CAO Rui, et al. Interaction mechanism and model of coal mine safety management behaviors[J]. China Safety Science Journal, 2025, 35(5): 32-38. doi: 10.16265/j.cnki.issn1003-3033.2025.05.1094
[2]	王秉, 周佳胜, 史志勇. 数智赋能安全态势感知与塑造模型[J]. 图书馆论坛, 2024, 44(8):1-7.
	WANG Bing, ZHOU Jiasheng, SHI Zhiyong. A model for safety & security situational awareness and shaping empowered by digital intelligence[J]. Library Tribune, 2024, 44(8): 1-7.
[3]	叶进杰, 李景升, 史志勇. 数智赋能情报驱动的关键信息基础设施安全韧性建设模型研究[J]. 情报杂志, 2025, 44(7):61-68,76.
	YE Jinjie, LI Jingsheng, SHI Zhiyong. Research on the model of safety resilience construction of critical information infrastructure driven by intelligence empowered by digital intelligence[J]. Journal of Intelligence, 2025, 44(7): 61-68,76.
[4]	JOHNSON D G, VERDICCHIO M. AI anxiety[J]. Journal of the Association for Information Science and Technology, 2017, 68:2 267-2 270. doi: 10.1002/asi.2017.68.issue-9
[5]	SUSENO Y, CHANG C, HUDÍK M, et al. Beliefs, anxiety and change readiness for artificial intelligence adoption among human resource managers: the moderating role of high-performance work systems[J]. The International Journal of Human Resource Management, 2021, 33:1 209-1 236. doi: 10.1080/09585192.2021.1931408
[6]	LI Jian, HUANG Jinsong. Dimensions of artificial intelligence anxiety based on the integrated fear acquisition theory[J]. Technology in Society, 2020, 63:DOI:10.1016/J.TECHSOC.2020.101410.
[7]	郭娟, 朱晓妹, 李姿颖, 等. 人工智能技术应用对员工心理状态的影响分析[J]. 中国人事科学, 2021(11): 50-58.
	GUO Juan, ZHU Xiaomei, LI Ziying, et al. Analysis on influence of application of artificial intelligence technology on psychological state of employees[J]. Chinese Personnel Science, 2021(11): 50-58.
[8]	刘小禹, 余彩婷. 悲欣交集:数字技术与员工情绪[J]. 外国经济与管理, 2024, 46(6): 134-152.
	LIU Xiaoyu, YU Caiting. A mixture of joy and sorrow: digital technologies and employee emotions[J]. Foreign Economics & Management, 2024, 46(6): 134-152.
[9]	杜佩佳, 刘生敏. 人工智能应用对员工心理的双刃剑影响[J]. 运筹与模糊学, 2024, 14(4): 541-547.
	DU Peijia, LIU Shengmin. The double-edged sword effect of artificial intelligence application on employees' psychology[J]. Operations Research and Fuzzy Systems, 2024, 14(4): 541-547.
[10]	蒋建武, 龙晗寰, 胡洁宇. 工作场所人工智能应用对员工影响的元分析[J]. 心理科学进展, 2024, 32(10):1 621-1 639.
	JIANG Jianwu, LONG Hanhuan, HU Jieyu. A meta-analysis of the effects of artificial intelligence application in the workplace on employees[J]. Advances in Psychological Science, 2024, 32(10): 1621-1639. doi: 10.3724/SP.J.1042.2024.01621
[11]	CHEN Min, NIKOLAIDIS S, SOH H, et al. Trust-aware decision making for human-robot collaboration[J]. ACM Transactions on Human-Robot Interaction (THRI), 2018, 9: 1-23.
[12]	胡保亮, 梅婕, 张素平. 人机协同决策:文献综述[J]. 信息与管理研究, 2024, 9(增1): 51-64.
	HU Baoliang, MEI Jie, ZHANG Suping. Human-AI collaborative decision-making: a literature review[J]. Journal of Information and Management, 2024, 9(Z1): 51-64.
[13]	田晓明, 段锦云, 孔瑜. 不确定管理理论:核心概念、内容介绍及未来展望[J]. 心理研究, 2012, 5(3): 70-77.
	TIAN Xiaoming, DUAN Jinyun, KONG Yu. A review of uncertainty management theory: conception, introduction and future directions[J]. Psychological Research, 2012, 5(3): 70-77.
[14]	BRASHERS D. Communication and uncertainty management[J]. Journal of Communication, 2001, 51: 477-497. doi: 10.1111/j.1460-2466.2001.tb02892.x
[15]	GARY Z. The influence mechanism of AI technology learning anxiety in the human-computer collaboration context: the moderating effect of uncertainty avoidance and the mediating effect of self-efficacy[J]. The EUrASEANs: Journal on Global Socio-Economic Dynamics, 2023, 6(43): 170-180.
[16]	SETHUMADHAVAN A. Trust in artificial intelligence[J]. Ergonomics in Design: the Quarterly of Human Factors Applications, 2018, 27:DOI:10.1177/1064804618818592.
[17]	罗方禄, 罗吴赞. 新质生产力安全发展的内涵及能力建设[J]. 中国安全科学学报, 2024, 34(10): 8-16. doi: 10.16265/j.cnki.issn1003-3033.2024.10.0616
	LUO Fanglu, LUO Wuzan. The connotation and capacity building of the safe development of new quality productive forces[J]. China Safety Science Journal, 2024, 34(10): 8-16. doi: 10.16265/j.cnki.issn1003-3033.2024.10.0616
[18]	MAYER R C, DAVIS J H, SCHOORMAN F D. An integrative model of organizational trust[J]. Academy of Management Review, 1995, 20(3): 709-734. doi: 10.2307/258792
[19]	MYERS C D, TINGLEY D. The influence of emotion on trust[J]. Political Analysis, 2016, 24: 492-500. doi: 10.1093/pan/mpw026
[20]	MATSUNAGA M. Uncertainty management, transformational leadership, and job performance in an AI-powered organizational context[J]. Communication Monographs, 2022, 89(1): 118-139. doi: 10.1080/03637751.2021.1952633
[21]	何贵兵, 陈诚, 何泽桐, 等. 智能组织中的人机协同决策:基于人机内部兼容性的研究探索[J]. 心理科学进展, 2022, 30(12):2 619-2 627.
	HE Guibing, CHEN Cheng, HE Zetong, et al. Human-AI collaborative decision-making in intelligent organizations: an exploratory study based on internal human-AI compatibility[J]. Advances in Psychological Science, 2022, 30(12): 2619-2627. doi: 10.3724/SP.J.1042.2022.02619
[22]	TOREINI E, AITKEN M, COOPAMOOTOO K, et al. The relationship between trust in AI and trustworthy machine learning technologies[C]. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency (FAT'20), 2020: 272-283.
[23]	KALTENBACH E, DOLGOV I. On the dual nature of transparency and reliability: rethinking factors that shape trust in automation[J]. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2017, 61: 308-312. doi: 10.1177/1541931213601558
[24]	胡艺龄, 陈彦君. 人机协作下的共享心智模型研究[J]. 现代教育技术, 2024, 34(1): 64-72.
	CHEN Yanjun, HU Yiling. Research on shared mental model under human-machine collaboration[J]. Modern Educational Technology, 2024, 34(1): 64-72.
[25]	BHASKARA A, SKINNER M, LOFT S. Agent transparency: a review of current theory and evidence[J]. IEEE Transactions on Human-Machine Systems, 2020, 50(3): 215-224. doi: 10.1109/THMS.6221037
[26]	GULATI S, SOUSA S, LAMAS D. Design, development and evaluation of a human-computer trust scale[J]. Behaviour & Information Technology, 2019, 38(10):1 004-1 015.
[27]	SHAMIM S, ZENG J, SHARIQ S M, KHAN Z. Role of big data management in enhancing big data decision-making capability and quality among Chinese firms: a dynamic capabilities view[J]. Information & Management, 2019, 56(6):DOI: 10.1016/j.im.2018.12.003.
[28]	POPE-DAVIS D, TWING J S. The effects of age, gender, and experience on measures of attitude regarding computers[J]. Computers in Human Behavior, 1991, 7: 333-339. doi: 10.1016/0747-5632(91)90020-2
[29]	MARGRETT J A, MARSIKE M. Gender differences in older adults' everyday cognitive collaboration[J]. International Journal of Behavioral Development, 2002, 26(1): 45-59. doi: 10.1080/01650250143000319 pmid: 20657668
[30]	INKPEN K, CHAPPIDI S, MALLARI K, et al. Advancing human-AI complementarity: the impact of user expertise and algorithmic tuning on joint decision making[J]. ACM Transactions on Computer-Human Interaction, 2022, 30: 1-29.
[31]	ZHANG Fenghua, XIAO Leifeng, GU Ruolei. Does gender matter in the relationship between anxiety and decision-making?[J]. Frontiers in Psychology, 2017, 8:DOI:10.3389/fpsyq.2017.02231.

模型	χ²	df	χ²/df	CFI	TLI	RMSEA	SRMR
4因子(J,X,T,C)	337.324	318	1.061	0.997	0.997	0.011	0.025 6
3因子(J,X+T,C)	1 669.131	321	5.200	0.806	0.788	0.090	0.076 3
2因子(J,X+T+C)	1 672.382	323	5.178	0.806	0.789	0.089	0.076 2
1因子(J+X+T+C)	1 673.218	324	5.164	0.806	0.789	0.089	0.076 1

变量	均值	标准差	相关系数
变量	均值	标准差	1	2	3	4	5	6	7	8	9
性别	1.48	0.50	1	—	—	—	—	—	—	—	—
年龄	2.5	0.90	-0.016	1	—	—	—	—	—	—	—
工龄	2.48	1.23	-0.056	0.751^#	1	—	—	—	—	—	—
学历	1.77	0.61	-0.025	0.007	0.038	1	—	—	—	—	—
行业	3.55	1.82	-0.019	-0.006	0.001	-0.045	1	—	—	—	—
AI焦虑	2.45	0.91	0.057	0.114^#	0.149^#	0.039	-0.008	1	—	—	—
人机信任	3.55	0.78	-0.069	-0.147^#	-0.173^#	-0.040	-0.005	-0.658^#	1	—	—
系统透明度	2.43	1.08	0.069	0.120^#	0.080	0.062	0.038	0.160^#	-0.174^#	1	—
人机协同决策质量	3.56	0.84	-0.094^*	-0.129^#	-0.156^#	-0.062	-0.050	-0.607^#	0.609^#	-0.133^#	1

变量类型	变量	人机信任		人机协同决策质量
变量类型	变量	模型1	模型2	模型3	模型4	模型5
控制变量	性别	-0.079	-0.023	-0.104^*	-0.052^*	-0.041^*
	年龄	-0.036	-0.034	-0.024	-0.022	-0.006
	工龄	-0.149^*	-0.021	-0.141^*	-0.021	-0.011
	学历	-0.036	-0.007	-0.062	-0.034	-0.031
	行业	-0.008	-0.013	-0.055	-0.059	-0.053^*
自变量	AI焦虑	—	-0.852^#*	—	-0.702^#*	-0.393^#*
中介变量	人机信任	—	—	—	—	0.480^#*
	决定系数R²	0.038	0.744	0.041	0.667	0.726
	解释力增量ΔR²	0.038	0.706	0.041	0.625	0.059
	统计量F	4.081^#*	249.559^#*	4.467^#*	172.165^#*	194.932^#*
	F变化量ΔF	4.081^#*	1 420.913^#*	4.467^#*	968.842^#*	111.107^#*

效应	效应值	标准误差	95%置信区间
间接效应	-0.337 1	0.035 7	[-0.451 2,-0.307 9]
直接效应	-0.362 1	0.041 8	[-0.444 2,-0.280 0]
总效应	-0.739 2	0.023 7	[-0.785 9,-0.692 6]

变量类型	变量	人机协同决策质量
变量类型	变量	模型6	模型7	模型8
控制变量	性别	-0.113^*	-0.050	-0.046
	年龄	0.003	0.025	0.021
	工龄	-0.163^*	-0.037	-0.035
	学历	-0.060	-0.032	-0.028
	行业	-0.046	-0.028	-0.024
中介变量	人机信任	—	0.817^#*	0.815^#*
调节变量	系统透明度	—	0.027	0.026
交互项	人机信任× 系统透明度	—	—	0.069^*
—	R²	0.042	0.683	0.687
—	ΔR²	0.042	0.641	0.005
—	F	3.806^#	132.775^#*	118.432^#*
—	ΔF	3.806^#	436.117^#*	6.403^#*