Making AI Work for the Public: An ALT Perspective

Based on two years of research—including more than 40 interviews, pilots in Boston, New York City, and San José, and a scan of national policy trends—we propose an alternative framework for public AI adoption: Adapt, Listen, and Trust (ALT).

Active

AI is rapidly reshaping the public sector, but most efforts remain focused on optimizing existing processes rather than reimagining how institutions serve communities. If governments continue to pursue efficiency alone, they risk entrenching the very systems that residents already distrust. Based on two years of research—including more than 40 interviews, pilots in Boston, New York City, and San José, and a scan of national policy trends—we propose an alternative framework for public AI adoption: Adapt, Listen, and Trust (ALT). Rather than reinforce the status quo, the ALT framework guides civic partners to build more responsive public institutions by (1) adapting to the amplified demand AI unleashes, (2) building shared civic infrastructure that enables genuine listening at scale, and (3) cultivating two-way accountability that deepens public trust. The report concludes by outlining concrete recommendations for governments, philanthropy, universities, and community organizations to align around the ALT approach.