AI is being deployed across a vast range of application domains, cultural environments, and human stakeholders. Safe and responsible AI development requires continuously accounting for diverse perspectives. We develop and test pluralistic agents, AI agent representatives that deliberate on behalf of individuals to align AI systems with diverse human perspectives.
Scaling public input on AI policy with pluralistic agents
calendar_today
November 13, 2026
domain
machinelibrary-ai