Artificial intelligence for educating family caregivers of people with dementia: A mixed methods systematic review

Dementia is a growing global health challenge, with the majority of individuals living with dementia receiving care from family members. However, many family caregivers lack sufficient knowledge, skills, and emotional support. Emerging artificial intelligence (AI) technologies present novel opportunities to deliver personalized, scalable, and accessible education for family caregivers. This mixed methods systematic review synthesized evidence on AI-based educational interventions for family caregivers of people with dementia. Nine electronic databases were searched for qualitative, quantitative, or mixed methods studies. Methodological quality was appraised using JBI tools and the Mixed Methods Appraisal Tool. Twelve studies met inclusion criteria (three qualitative, two quantitative, seven mixed methods). The AI interventions employed included chatbots, virtual assistants, systems based on large language models (LLMs), AI-driven decision-support tools, and adaptive VR simulations capable of dynamically responding to user input. Preliminary evidence suggests potential improvements in caregiver knowledge, digital literacy, emotional well-being, and caregiving competencies. However, the outcomes assessed and reported effects were heterogeneous across studies. Overall, the AI-based interventions were perceived as acceptable and feasible. However, the current evidence base remains limited in scope and methodological rigor, with few randomized controlled trials and considerable variability in study designs and AI modalities. AI technologies hold promises for enhancing dementia caregiver education. Future research should prioritize long-term outcomes, cultural relevance, and ethical considerations to support scalable, person-centered interventions.