Insight

Building integrated government systems with AI

To build human-centered systems, public servants should orchestrate a dynamic interplay between AI tools, previously-siloed agency systems and workflows, and community needs.

PublishedJune 2, 2025

Authors

As artificial intelligence (AI) becomes embedded in public service delivery, a quiet but profound paradigm shift is underway. The future of AI in civic tech isn’t just about smarter tools, but about reimagining roles, responsibilities, and relationships.

Traditionally, public servants — especially front-line workers — have acted as expert guides through often-fragmented systems. They’ve been the “keepers of the map,” helping people navigate complicated benefit applications, verify income, and get through other bureaucratic hurdles. But now, as AI tools can help draft application forms, flag missing documentation, and even offer interpretations of complex policies, the role of human expertise is transforming.

We think a better metaphor for this new role is not navigator or builder, but choreographer.

In this emerging paradigm, public servants are not getting replaced by AI but are orchestrating a dynamic interplay between AI tools, previously-siloed agency systems and workflows, and community needs.

Our piloting to date shows that a caseworker doesn’t hand over their job to a chatbot — they decide how and when a chatbot can support their work.

A policy expert doesn’t simply review outputs — they shape the rules that are encoded into AI tools, ensuring they remain fair and accurate.

A designer doesn’t just make screens — they choreograph how AI, human action, and user input come together across a service journey.

This shift is happening against a backdrop of real pressures and real risks. In the face of budget constraints, some governments may reach for AI as a blunt cost-cutting tool, aiming to replace humans wholesale. That path can lead to brittle systems and harmful mistakes.

But it doesn’t have to be this way. When the introduction of AI capabilities takes appropriateness seriously — such as applying the steps that our team published — and is paired with rigorous evaluation and human-centered design, it can make public services faster, simpler, and more responsive.

The future we envision is one of human-centered systems integration: where teams of designers, engineers, data scientists, and front-line staff come together to choreograph AI’s role in public services. It’s a future that demands more expertise, not less, and one where success is measured not just by automation, but by outcomes and trust.

This post is an excerpt from the first of a new Nava Labs Newsletter. If you'd like to sign up for our Nava Labs newsletter, you can do so here.

Written by


Martelle Esposito

Partnerships and evaluation manager

Martelle Esposito is a partnerships and evaluation lead at Nava. Before joining Nava, Martelle managed a WIC services innovation lab at Johns Hopkins University and worked on public policy development and program implementation at non-profits.

Genevieve Gaudet

Director of Design

Genevieve Gaudet is the Director of Design, Growth and Strategy at Nava. Before joining Nava, Genevieve co-founded the NYC Service Design Studio in the office of Mayor Bill de Blasio.

Partner with us

Let’s talk about what we can build together.

Get in touch