Artificial Intelligence and Mental Health Help-Seeking in Asia: A Scoping Review

Team: Brendon Ban Ren Jie (PI, Master's student, [email protected]); Assoc. Prof. Dr Lau Poh Li (supervisor); two research assistants (RAs)

Design: JBI scoping review reported with PRISMA-ScR; protocol registered on OSF; October 2026 to February 2027

Background and Rationale

Help-seeking is typically conceptualised as a staged process of recognising a difficulty, appraising it as warranting support and selecting a source of help (Rickwood et al., 2005). In Asia, stigma, concern for face and familial obligation constrain each stage, and among men, conformity to masculine norms is associated with poorer mental health and less favourable attitudes towards psychological help-seeking (Wong et al., 2017). AI conversational agents reduce symptoms of depression and distress (Li et al., 2023), yet heavier chatbot use has been linked to greater loneliness and reduced socialisation (Fang et al., 2025). Whether AI acts as a bridge to professional care or a substitute for it therefore remains unresolved. Because this literature is dispersed across disciplines, inconsistently operationalised and drawn largely from Western samples, a scoping review is the appropriate first step (Munn et al., 2018).

Aim and Review Questions

Aim: To map research on AI and mental health help-seeking in Asia, with particular attention to men, and to identify gaps. Questions: What is known about AI in relation to mental health help-seeking in Asia, specifically (a) the AI types, populations and countries studied, (b) the help-seeking stages addressed and how they are measured, and (c) whether AI is positioned as a bridge to or a substitute for human support?

Eligibility Criteria

Population: People in Asia or of Asian background; any age or gender
Concept: Any AI application examined in relation to help-seeking attitudes, intentions, behaviour or pathways to care
Context: Any Asian country (UN geoscheme) and any setting
Sources: English-language peer-reviewed journal articles, 2015 to 2026; theses and grey literature excluded

The 2015 start date reflects the period in which conversational AI became widely available to the public; criteria will be refined, if necessary, after piloting the search and documented as protocol amendments.

Methods and Responsibilities

Stage	Procedure	Who
Search	PubMed, PsycINFO, Scopus, Web of Science, ACM, IEEE Xplore, MyCite, ASEAN Citation Index; reference lists	PI; supervisor reviews strategy
Screen	Pilot calibration, then independent dual screening of titles, abstracts and full texts	PI screens all; RAs share second screening
Chart	Piloted form: study details, country, sample (incl. men), AI type, help-seeking outcomes, cultural issues	RAs chart; PI verifies 20%
Synthesise	Counts, evidence map and narrative by help-seeking stage; no quality appraisal	PI
Report	PRISMA-ScR checklist and flow diagram	PI and supervisor

The review follows JBI guidance (Peters et al., 2020) and PRISMA-ScR (Tricco et al., 2018). Disagreements are resolved by discussion or by the supervisor.

Timeline

Review timeline · 7 activities over 4 months

Significance and Outputs