Generative AI · Education leadership

Actionable companion · Governance, equity and accountability

From evidence to district action: AI chatbots for health information-seeking and self-diagnosis among Saudi university students: patterns of use, predictors, attitudes, and trust

AI-assisted editorial implementation guide. The district actions are proposed applications, not findings or recommendations attributed to the original authors. No additional peer review is asserted.

Read original paperSearch paper agentCC BY 4.0

Evidence and context

Confirm the study population, setting and task before applying its findings to your district.

Original authors’ abstract, reproduced under the linked source license:

Background The diffusion of large-language-model (LLM) chatbots such as ChatGPT has introduced a new layer to the digital health information ecosystem, alongside search engines, social media, and national mobile-health (mHealth) platforms. The actual prevalence of AI chatbot-based self-diagnosis among Saudi university students, the ecosystem of tools they choose, and the factors that predict use are not well characterised. Objectives To estimate AI chatbot use for self-diagnosis among Saudi university students, map the ecosystem of tools used, characterise attitudes and perceived superiority over traditional search, identify predictors through logistic regression, and surface qualitative themes. Methods A cross-sectional online survey was administered to 511 Saudi university students between April and May 2024. Instruments included a binary item on prior self-diagnosis use, a multi-response item on AI tools used, an 8-item AI Chatbot Perceived Superiority scale, a 15-item AI Doctor Chatbot Attitudes and Trust scale, five multi-response items, and three open-ended items analysed thematically. Data were analysed using descriptive statistics, multivariable logistic regression, Mann–Whitney U/Kruskal-Wallis tests, tool co-use analysis, and latent class analysis. The study received IRB approval from Qassim University (No. 23-19-02). Results ChatGPT was the dominant tool (n = 355), followed by Google Gemini (n = 197), Microsoft Copilot (n = 166), and the Ministry of Health’s Sehhaty AI chatbot (n = 154). Multivariable logistic regression identified three independent predictors: gender [female vs. male, adjusted OR = 0.32, 95% CI (0.20, 0.52), p < 0.001], age (per step OR = 1.24, p = 0.080), and specialty (Health vs. Education OR = 6.71, p < 0.001). AI was rated easier and faster than search engines (M = 3.51) and more comprehensive (M = 3.43). Privacy concerns were prominent (M = 3.69); 65.1% endorsed institutionally verified AI chatbots. Tool co-use analysis showed the seven AI tools were used largely independently (max phi = 0.12). Conclusion AI doctor chatbots have already become mainstream tools for self-diagnosis among Saudi university students, with ChatGPT dominating but the Sehhaty AI chatbot already in active use. Male students and those in Health, Management, Social Science, and Engineering are substantially more likely to use AI. The strong endorsement of institutionally verified chatbots provides a clear policy signal for embedding safe, MOH-aligned LLM tools within Sehhaty under Vision 2030.

A 90-day district pilot

  1. Days 1–15 · Appraise. Have the curriculum lead and a teacher read the methods and limitations. Record the study design, sample, comparator, outcomes and evidence gaps. Define a local question that the study can help inform.
  2. Days 16–30 · Design. Convene a teacher, student, family and technology working group. Test one proposed use against a documented privacy, accessibility, bias and human-review checklist before authorizing a pilot. Name an accountable owner and set the baseline, success threshold and budget before implementation.
  3. Days 31–60 · Test. Start with a small voluntary cohort, preserve an accessible non-AI route and require teacher review. Use synthetic or de-identified examples; approve data handling and tool terms through district processes.
  4. Days 61–90 · Decide. Track unresolved risks, subgroup access, complaint resolution time and the proportion of outputs that require substantive human correction. Compare with baseline, document uncertainty and stop or revise if learning worsens, access is unequal or critical errors persist.

The board decision brief

Report the learning problem, evidence quality, participating population, total cost including review time, measured benefit, errors, access differences and remaining uncertainties. Make a dated recommendation to stop, revise or expand. A positive perception survey alone does not establish learning improvement.

Attribution and reuse

Haitham Alzghaibi; Manahil Alfuraydan. AI chatbots for health information-seeking and self-diagnosis among Saudi university students: patterns of use, predictors, attitudes, and trust. Frontiers in Public Health (2026-10-08). https://doi.org/10.3389/fpubh.2026.1899249

Companion prepared 2026-10-30. Original authorship and DOI are preserved. The abstract is an attributed source extract; the implementation plan is new AI-assisted editorial material, offered under the same CC BY 4.0 license. This guide has not undergone independent peer review.