Privacy Concern, Transparency Engagement, and Public Accountability Expectations in AI Platform Governance: Evidence from Tourism Services

This secondary analysis examines factors associated with user accountability expectations in AI platform governance, using tourism services as the empirical setting. The study draws on 311 valid survey responses collected after ethical approval and geolocated to Yangsan, Het Hogeland, and Paris. The three groups are treated as local convenience samples rather than representative samples of their cities or countries. Descriptive comparisons, nonparametric tests, and binary logistic regression with HC3 robust standard errors were used to examine privacy concern, transparency engagement, AI service use breadth, and geographic sample membership. Privacy concern showed the strongest adjusted association with support for public disclosure responsibility, OR = 5.39, 95% CI [2.56, 11.35]. Respondents who actively read notices about platform data use or privacy also had higher adjusted odds of supporting public disclosure responsibility, OR = 2.93, 95% CI [1.09, 7.86]. AI service use breadth and geographic sample membership were not significant independent factors. Privacy concern differed modestly across the three samples, while accountability response patterns were broadly similar. The findings suggest that stronger accountability expectations were associated more closely with privacy concern and attention to transparency information than with service exposure or geographic context. Functional use of AI services did not appear to correspond directly to stronger awareness of public accountability. These results provide preliminary empirical evidence for user facing accountability design in AI systems, including understandable explanations, accessible appeal mechanisms, and routes to human review when automated outcomes are disputed.隐私担忧与支持公开披露责任之间存在最强的调整后关联，OR = 5.39，95% CI [2.56, 11.35]。积极阅读平台数据使用或隐私相关通知的受访者，其支持公开披露责任的调整后几率也更高，OR = 2.93，95% CI [1.09, 7.86]。AI服务使用广度和地理样本成员身份并非显著独立影响因素。三个样本中隐私担忧存在轻微差异，而问责回应模式总体相似。研究结果表明，更强的问责预期与隐私担忧以及对透明信息的关注关系更密切，而非与服务接触程度或地理背景相关。AI服务的功能性使用似乎并未直接对应于对公共问责性的更高认知。这些结果为人工智能系统中面向用户的问责设计提供了初步的实证依据，包括提供可理解的解释、便捷的申诉机制，以及在自动化结果受到质疑时提供人工复核的途径。