MindBridge: Responsible AI Chatbots for Graduate Student Mental Health Support in Higher Education

MindBridge

Fahima Tabassum, Muhammad Abu Naim, Saida Taharima

Research (anything related to the use of AI in a research context)

Our team investigates the use of AI chatbots as a mental health support tool for graduate students in higher education. Graduate students often face significant psychological pressure related to research progress, career uncertainty, financial stress, and social isolation. These challenges can be particularly intense for international students who must also navigate cultural and language barriers. In our project, we explore how an AI chatbot could provide accessible, low-barrier emotional support and guidance for graduate students. The situation we study involves students interacting with a conversational AI system during moments of stress related to their research life or personal concerns. The participants include graduate students (master’s and PhD), both domestic and international, studying at universities in Japan. To simulate this use case, we conducted a survey and an experimental interaction with an AI chatbot. Participants first completed a Big Five personality assessment and then consulted an AI system about concerns related to their academic life, such as research progress, financial worries, and personal relationships. They then evaluated the AI’s responses in terms of usefulness, empathy, and overall satisfaction. Based on these interactions, our team is developing a concept for an AI mentoring application called “Dr. Bridge.” The goal is to personalize AI responses according to users’ personality traits and needs, creating a more supportive and context-aware system. The project also examines ethical considerations such as privacy, emotional safety, and the limits of AI in providing psychological support.This use case allows us to analyze both the potential benefits and risks of AI-assisted mental health support in higher education, and to explore how such systems could complement existing university counseling services while maintaining responsible and ethical AI use.

This use case is important because mental health challenges among graduate students are becoming increasingly visible in higher education systems worldwide. Research pressure, uncertainty about future careers, financial concerns, and social isolation can significantly affect students’ well-being, academic performance, and long-term professional development. For international students, these pressures may be amplified by cultural adjustment, language barriers, and limited local support networks. Artificial intelligence offers a potential opportunity to provide accessible and low-barrier support systems for students who may hesitate to seek traditional counseling services. AI chatbots can provide immediate responses, offer a private space for reflection, and be available at any time, which may help students manage stress and organize their thoughts about academic and personal challenges. Such tools could serve as an early support mechanism, encouraging students to reflect on their concerns and potentially guiding them toward appropriate support resources. However, the use of AI for mental health support also raises important ethical and governance questions. Issues such as data privacy, emotional safety, cultural sensitivity, and the risk of over-reliance on automated systems must be carefully considered. International organizations such as the OECD and the Global Partnership on AI have emphasized the importance of responsible and human-centered AI systems, particularly when technologies interact with vulnerable individuals. By studying this use case, our project aims to better understand how AI-based tools can responsibly complement existing university counseling services while ensuring that technological innovation supports student well-being without replacing essential human care. This work contributes to broader discussions on how universities and policymakers can integrate AI into higher education in a way that is ethical, inclusive, and beneficial for diverse student communities.

Our team decided to participate in the AI Grand Challenge because the topic we are studying is closely connected to our own experiences as graduate students. During our academic journeys, we have seen how research pressure, uncertainty about the future, financial concerns, and feelings of isolation can affect students’ mental well-being. For international students in particular, adapting to a new culture and academic environment can make these challenges even more complex. These experiences motivated us to explore whether artificial intelligence could play a supportive role in helping students manage stress and seek guidance when they need it. Through this project, we want to better understand how people interact with AI systems when discussing personal or emotional concerns. We are especially interested in whether conversational AI can provide meaningful support, how personalization—such as using personality traits—might improve the user experience, and how students from different cultural backgrounds perceive and trust AI responses. At the same time, we recognize that using AI for mental health support raises important ethical questions. We want to learn more about how issues like privacy, emotional safety, transparency, and responsible AI governance should be addressed when developing such tools. Many international discussions, including those led by organizations like the OECD and the Global Partnership on AI, emphasize that AI systems should remain human-centered and trustworthy. By taking part in this challenge, we hope to learn from experts, mentors, and other student teams around the world. More importantly, we want to contribute a student perspective to the conversation about how AI can be responsibly integrated into higher education in ways that support student well-being while respecting ethical boundaries.

1. Situation / Context Graduate students worldwide increasingly experience significant psychological pressure throughout their academic journey. Research deadlines, publication expectations, career uncertainty, financial stress, and social isolation can create a demanding environment that negatively affects both well-being and research productivity. These pressures are often intensified for international students, who must additionally navigate cultural adaptation, language barriers, and distance from family support networks. Although universities have expanded counseling services in recent years, many institutions still face limitations such as long waiting times, limited counselor availability, and barriers related to stigma or accessibility. As a result, some students hesitate to seek professional help even when they experience considerable stress. At the same time, rapid advances in artificial intelligence have introduced new possibilities for digital support systems. Conversational AI chatbots are increasingly being used in educational and healthcare settings to provide guidance, information, and emotional support. Because AI chatbots are available at any time and can provide immediate responses, they may offer a low-barrier entry point for students who need support but hesitate to approach traditional counseling services. In this project, our team investigates the potential role of AI chatbots as a complementary mental health support tool for graduate students in higher education. To explore this use case, we conducted an exploratory study involving graduate students studying in Japan, including both domestic and international participants. Participants first completed a Big Five personality assessment, which evaluates five major personality dimensions: openness, conscientiousness, extraversion, agreeableness, and neuroticism. After the assessment, participants interacted with an AI chatbot and discussed concerns related to academic life, such as research progress anxiety, financial pressure, and personal relationships. Following the interaction, participants evaluated the AI responses in terms of usefulness, empathy, and overall satisfaction. In total, 42 graduate students participated in this exploratory investigation, providing both quantitative and qualitative feedback on their experience with AI-assisted consultation. Based on these observations, our team proposes the concept of a personalized AI mentoring system called “Dr. Bridge.” This system aims to provide supportive and context-aware responses tailored to individual personality traits while maintaining ethical safeguards for responsible AI use. 2. Critical Analysis of the Situation Our investigation highlights both the opportunities and the risks associated with the use of AI chatbots in student mental health support. One major advantage of AI systems is accessibility. AI chatbots can provide immediate responses at any time, which may help students express their concerns during moments of stress. For some students, interacting with an AI system may feel less intimidating than speaking directly with a counselor, especially when discussing sensitive personal issues. However, several critical limitations must be considered. First, AI systems do not possess genuine emotional understanding. Although they may generate responses that appear empathetic, their responses are produced through pattern recognition rather than real emotional experience. As a result, AI may struggle to respond appropriately to complex psychological situations. Second, the use of AI for mental health support raises important ethical concerns regarding privacy and data governance. Conversations related to emotional well-being involve highly sensitive personal information. Universities must therefore ensure that any AI-based system operates under transparent policies regarding data security, informed consent, and responsible data use. Third, cultural diversity presents an additional challenge. Graduate student populations are increasingly international and culturally diverse. AI systems trained on limited datasets may fail to recognize cultural nuances, potentially limiting their effectiveness for international students. Finally, there is a broader risk of over-reliance on automated systems. AI tools should not replace professional counseling services. Without proper safeguards, students may depend too heavily on AI systems rather than seeking support from trained mental health professionals. These concerns highlight the need for responsible AI governance in higher education, consistent with international principles such as the **OECD AI Principles and the policy discussions promoted by the Global Partnership on AI. These frameworks emphasize transparency, accountability, human-centered design, and the protection of fundamental rights in AI deployment. 3. Perspectives Discussed Within Our Team Throughout this project, our interdisciplinary team engaged in extensive discussions about the potential role of AI in supporting graduate student well-being. One perspective emphasized the opportunity to expand access to support services. AI chatbots could function as an initial point of contact where students can express concerns, organize their thoughts, and receive guidance before seeking professional counseling. Another perspective emphasized the limitations of AI systems in emotionally complex contexts. Some team members argued that mental health support fundamentally requires human empathy, contextual understanding, and professional expertise that AI systems cannot fully replicate. Our team also discussed ethical responsibilities associated with introducing AI tools into sensitive areas such as mental health support. Key issues included transparency about AI-generated responses, informed consent from users, and the need for clear pathways to escalate serious concerns to human professionals. Through these discussions, we reached a shared conclusion: AI systems should be designed as supportive tools that complement human counseling services rather than replace them.