“I talk to the chatbot more than I talk to anyone.”
Does talking to an AI companion lead Malaysian men toward help from friends, family and counselors, or away from it? We think self-compassion is the key.
AI chatbots now give Malaysian men something they have rarely had: a private, judgment-free place to talk at any hour. Nobody yet knows whether that leads them toward people or away from them.
Malaysian men are less likely than women to seek help for psychological distress, and many wait until they reach a crisis. Conformity to masculine norms, particularly self-reliance, is consistently linked to poorer mental health and more negative attitudes toward seeking help (Addis & Mahalik, 2003; Wong et al., 2017).
In Malaysia these norms are reinforced by concern for face. For many men, admitting they are struggling feels like failing.
Noticing that something is wrong.
Judging that it warrants support.
Choosing who or what to turn to.
Rickwood et al., 2005. These models were built before AI was a realistic source of support.
Conversational agents reduce depressive symptoms, and brief chatbot conversations have increased self-compassion.
Li et al., 2023 · Lee et al., 2019
In a four-week randomized study by MIT Media Lab and OpenAI, heavier chatbot use was linked to loneliness and less contact with other people.
Fang et al., 2025
Self-compassion is the capacity to treat yourself with kindness during difficulty (Neff, 2003). By buffering the shame that masculine norms attach to asking for help, it may make men more willing to reach out. Heath et al. (2017) found exactly this among men: the link between masculine norms and self-stigma was weaker for those higher in self-compassion.
The program draws on the Theory of Planned Behavior (Ajzen, 1991) and tests whether an AI companion increases actual help-seeking from people, and why.
H1 is tested in Study 2; H2 to H4 are estimated in Study 2 and tested in Study 3.
Universities and community services gain a low-cost, culturally fitted way to reach men who would never walk into a counseling room.
Developers and regulators gain clear evidence that AI tools must be designed to hand users over to human support.
This builds on our earlier survey of 210 Malaysian undergraduates, where social connection (r = .43) and empathy (r = .53) were linked to willingness to seek help (Chong, Ng, Ban et al., 2022).
Each study answers one question and feeds the next. Study 3 goes ahead only if Study 2 meets its progression criteria.
Purple marks the decision gate in August 2027 and final publications in 2029.
Ora is the study app. Sol is the AI companion inside it. Both already exist as a working prototype, with a safety escalation protocol and a researcher console, so the trials can begin as soon as ethics approval is granted. Every safety flag is followed up by a person within 24 hours.
We are mapping published research on AI and mental health help-seeking in Asia, with particular attention to men. The gaps we find will shape the design of Sol and the outcomes we measure in Studies 2 and 3.
A small randomized trial of Ora and Sol with young Malaysian men. It tests whether a full trial is workable and safe, and gives a first look at whether Sol works as a bridge or a substitute. Findings will be written up as manuscripts for publication.
A larger trial across several Malaysian settings to give a clear answer on whether Sol helps men reach people, and whether self-compassion explains the effect. It goes ahead only if Study 2 shows a full trial is workable.
Want to get involved in Bridge or Substitute? Great. We are building a small team across all three studies: research assistants for the scoping review and the two trials, and people who can code and work with AI to build Ora and Sol. All positions are unpaid and voluntary, and we are open to collaboration. This page explains what is open right now and what each role involves.
The summary as of today. Read below for the details.
Before you apply, ask yourself what you want from the experience. For some people it is a step toward a Master's or PhD; for others it is a way to find out whether research, or building technology for mental health, suits them at all. The scoping review is a structured place to learn how evidence is gathered. The trials show how a study runs with real participants. The tech work puts an AI companion in front of the men it is meant to help. It helps to come with a clear idea of why this project, and why now.
If after reading this entire page you'd like to be considered, email 22059169@siswa.um.edu.my with the role in the subject line. The Study 1 position closes October 30, 2026; the trial and tech roles are open on a rolling basis.
Work as a second reviewer alongside the principal investigator and one other research assistant. Hands-on research experience in digital mental health, for students and recent graduates.
Attach your resume or CV (academic background and any research experience) and your most recent academic transcript. In the email body, tell us briefly why you are interested, what relevant experience or skills you have, and what you hope to gain.
22059169@siswa.um.edu.myHelp run the feasibility trial with 80 Malaysian men from January 2027, and the full multi-site trial that may follow from 2028. Good preparation for anyone heading toward clinical or health research.
Ora is the study app and Sol is the AI companion inside it. Both exist as a working prototype. We need people who enjoy building things to help make them better and safer before and during the trials.
Email your CV and a short note on why you are interested and what you would bring. For tech roles, include a link to your GitHub or portfolio if you have one. These roles are unpaid and voluntary. We will agree hours, recognition and authorship credit with you before you start. Collaborators are welcome to use the same address.
Papers, protocols and talks from the program, including work in preparation and planned outputs. Filter by type or status, or tap a hashtag.
Titles of planned work are working titles and may change.
Short posts on what we are doing, why, and what we are learning along the way.
Change anything on the site. Use the tabs for lists like the team, posts and timelines, and Edit page text to change any heading, paragraph or table cell in place. Changes stay as a draft until you publish them.
Meet the people behind Bridge or Substitute, a three-study program running from 2026 to 2029 in the Department of Educational Psychology and Counseling, Universiti Malaya.
Master of Professional Counseling, Universiti Malaya
Brendon is a counseling trainee who understands both the clinical side and the technology. He designed and built Ora and Sol, including the safety escalation protocol and the researcher console, so the trials can start as soon as ethics approval is granted. He leads all three studies: he runs the Study 1 search and screening, and he is writing up each study as a manuscript for publication.
His earlier work, a survey of 210 Malaysian undergraduates presented at MAPSA 2022, found that social connection and empathy were linked to willingness to seek help. The program builds on that finding.
Department of Educational Psychology and Counseling
Dr. Lau supervises the program. She reviews the Study 1 search strategy and settles screening disagreements the team cannot resolve by discussion. She also oversees ethics approval with the Universiti Malaya Research Ethics Committee and the conduct of the trials in Studies 2 and 3.
Study 1 · two positions · recruiting
Two research assistants work as second reviewers on the scoping review. They share the screening of titles, abstracts and full texts, and chart the included studies in a shared Excel form. They are trained in JBI and PRISMA-ScR methods and named in the published article. This could be you.
Studies 2 and 3 · open · rolling applications
Research assistants for the two trials, and people who code and work with AI to build and improve Ora and Sol. If you are interested in programming, AI or running a trial, we would like to hear from you.