Journal of Technology in Behavioral Science Health (social science)

The CARE Framework: A Psychologically Grounded and Ethically Governed Approach to Culturally Adaptive and Emotionally Intelligent Artificial Intelligence

Mirza Jahanzeb Beg, Shrijitha S., Nivetha S., Mirza Zamin Raja Beg, Shahnaz Shah, Seshadri Sekhar Chatterjee
Springer Nature · 2026-10-21 · DOI 10.1007/s41347-026-00735-y

Abstract

Abstract Emotionally intelligent artificial intelligence (AI) is increasingly used in healthcare, education, and social robotics, yet most systems treat emotion as a fixed signal read from the face, voice, or text. This approach assumes universality, separates affect from cognition, and treats ethics as an afterthought. To propose CARE, a conceptual framework for designing and evaluating emotionally intelligent AI around four pillars: Cultural Adaptivity, Affective–Cognitive Integration, Regulation and Resilience, and Ethical Alignment. An integrative critical review of affective neuroscience, social cognition, emotion regulation, cultural psychology, and responsible AI was used to derive four design requirements, each linked to measurable constructs and candidate instruments. No single pillar is new. CARE’s claim is that the four are interdependent, so addressing any one in isolation produces predictable failures; we state this claim as six falsifiable propositions. CARE treats emotional meaning as context-dependent rather than as a stable property awaiting detection, a position consistent with the European Union AI Act’s prohibition on inferring emotion from biometric data in workplaces and education. Emotionally intelligent AI is as much a psychological, cultural, and ethical undertaking as an engineering one. CARE offers a testable scaffold and validation agenda of particular relevance to the Global South, where systems trained on Western data are likely to perform worst and evidence is scarcest.

Read at publisher →