Humane Intelligence in Geropsychiatric Care: Relational Artificial Intelligence, Clinical Wisdom, and the Moral Grid...

Artificial intelligence (AI) already influences how older adults are identified for services, supported between provider visits, and referred for care, yet most AI governance currently focuses on algorithms and infrastructure rather than the actual experiences of older adults and their caregivers. Humane Intelligence is a patient-centered, ethically attuned, relational framework for designing, evaluating, and monitoring AI in older adult care. It rests on four pillars: Relational Intelligence, Transparency with Care, Reciprocity and Consent, and Ethical Governance in Strategic Regions, and applies them from point-of-care encounters to system-level decisions. For each pillar, guidelines follow the same sequence: signal a problem, take action, and verify benefit or harm. The framework prioritizes outcomes that matter in geriatric psychiatry, including function, distress, caregiver burden, avoidable utilization, equity, and documented harms or overrides. To protect patients, it draws a firm boundary against fully automated clinical actions and recommends clinical-grade standards for patient-facing programs, including scope-of-action labels, human-in-the-loop safeguards for high-risk situations, postmarket monitoring, and periodic certification. This blueprint aligns with 2025 JAMA Summit on AI priorities, World Health Organization guidance for large multimodal models, United States Food and Drug Administration recommendations for Predetermined Change Control Plans, and Office of the National Coordinator for Health Information Technology decision-support intervention frameworks. The goal is practical: to translate ethics into testable patient-centered routines that clinicians can trust, healthcare systems can implement, and leaders can procure, so that AI augments rather than displaces care.