Into the Unknown: Leveraging Conversational AI in Supporting Young Migrants' Journeys Towards Cultural Adaptation

Research Background and Issues

• Identified Problems or Challenges: The paper highlights that accelerated globalization has made migration a common phenomenon, with young migrants facing significant challenges during the cultural adaptation process. These challenges include language acquisition, building social relationships, and balancing cultural identity between their home culture and the host country's culture. Furthermore, existing conversational AI systems primarily focus on administrative tasks and have yet to adequately support cultural adaptation.
• Importance of the Issue: Young migrants are at a critical stage of cultural identity development. Successful cultural adaptation is not only crucial for personal growth but also positively impacts the overall social integration and development of migrant communities. Providing targeted support to young migrants can alleviate cultural conflicts within families and promote social and economic integration.
• Research Motivation and Related Work: Although HCI research has explored technologies to support migrants' cultural adaptation, existing systems often adopt a one-size-fits-all approach, lacking support for personalized cultural adaptation processes. Previous solutions have largely focused on language learning or administrative support for migrants (e.g., translation, access to health information, and employment guidance), with insufficient attention paid to the dynamic process of cultural identity formation.

Solution

• Proposed Method or Solution: Through interviews and workshops with individual migrants and relevant stakeholders (including AI developers, policy experts, and social scientists), the authors explore the potential of applying large language model (LLM)-based conversational AI to support young migrants in their cultural adaptation. The study proposes that AI can serve as a guiding tool during the initial stages of cultural adaptation, assist migrants in reflecting on their culture during the intermediate stages, and help balance and integrate home and host cultures during the later stages of adaptation.
• Innovations:
  • Designing culturally sensitive AI to support personalized cultural exploration and identity regulation.
  • Proposing multimodal functionalities (e.g., image generation) and dialogue-driven interactions to help migrants construct more concrete and immersive cultural adaptation scenarios.
  • Emphasizing a dual perspective that balances individual and societal needs from both the migrant and host society viewpoints.
• Implementation Steps:
  • Phase 1: Conduct workshops with 11 young migrants, using journey mapping to analyze their cultural adaptation processes and guide the design of personalized ChatGPT systems.
  • Phase 2: Conduct focus group interviews with 9 stakeholders to explore opportunities and challenges for practical applications.
  • Data analysis employs thematic coding and affinity diagramming to integrate the perspectives and needs of migrants and stakeholders.

Research Findings

• Specific Findings:
  • Identified how young migrants expect conversational AI to support them at different stages of adaptation (initial, intermediate, and later stages), such as helping them understand host country social norms, reflect on their home culture, and integrate the two cultures.
  • Determined design directions for conversational AI in cultural adaptation, including avoiding cultural biases, supporting open-ended dialogue, and offering diverse options to enhance user autonomy.
• Advantages: Compared to existing solutions, this study emphasizes personalized support, advances culturally sensitive technology, and integrates migrant needs with host country policy requirements.
• Experimental or Evaluation Results:
  • Participants confirmed that the multimodal capabilities of conversational AI (e.g., image generation) helped them imagine and prepare for cultural adaptation scenarios, although some results were overly idealized.
  • Stakeholders suggested using data analysis to design more inclusive migration policies and community activities, fostering mutual integration between migrants and society.
• Limitations and Future Directions:
  • Limitations: The study primarily involved Nordic young migrants with strong digital skills, which may not fully represent the needs of other migrant groups; technical implementation details remain unspecified; the long-term dynamics of cultural adaptation were not comprehensively captured.
  • Future Directions: Expanding the geographical and skill diversity of participants, researching the technical implementation of culturally sensitive AI, exploring long-term user-AI interaction during adaptation, and developing integrated solutions with community support systems.

Conclusion

By integrating the perspectives of migrants and stakeholders, this study provides a theoretical and practical framework for applying LLM-based conversational AI in cultural adaptation. Despite existing challenges, the research suggests empowering migrants to define their cultural identity and navigate their adaptation journey through open-ended interactions and culturally sensitive design. It also emphasizes the importance of mutual integration with society for the successful implementation of this technology.