Long-term preservation of human agency in AI-mediated institutions for low-income communities
open
United States, North America
WS04403
As AI systems become embedded in education, employment, finance, healthcare, and government, people may technically retain choices while practical options are shaped by invisible optimization systems. A useful solution must preserve meaningful consent, contestability, and human responsibility. The solution must work with limited capital, intermittent services, and local maintenance capacity. The research opportunity is to define measurable success criteria, test the approach in realistic settings, identify failure modes, and create a pathway from prototype to accountable deployment.