IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

In the Age of AI, a New Approach to Public Engagement

California is supplementing its traditional process of navigating public comments with AI, to do a better job of gathering actionable insight. Its model may inform similar engagement efforts by other governments.

A person's hands use a keyboard in blue light, while white figures of people stand on the screen.
Shutterstock
The state of California is gathering statewide input about AI technology — and using AI to turn that input into policy recommendations that could be implemented, a public engagement approach which could be an example for other governments.

This is the third project using the Office of Data and Innovation (ODI)’s digital conversation platform, Engaged California, which was designed to enable discussions that could shape public policy. Its first pilot, in February 2025, focused on aiding victims of the L.A. wildfires that January. The second pilot involved gathering input from state workers about government efficiency.

The latest project centers on the California public to help officials understand how AI is affecting work. The feedback opportunity is open to every California resident to share their thoughts. The questions reflect areas in which many states are seeking information: How has AI impacted your job and workplace? How do you expect AI to impact the economy? What do you think the government should do about these impacts?

Engaged California is not intended to replace current engagement methods, but rather, to supplement them, according to ODI Director Jeffery Marino, acting as a sort of “digital town hall” to make public engagement more accessible and equitable.

“I would welcome any and all perspectives,” Jarrett Krumrei, ODI’s lead product manager, said, underlining the significance of people participating in this initiative as AI is impacting jobs across industries.

AI has two roles in the Engaged California methodology. First, a large language model (LLM) empowers the team to ask questions about the comments in plain language, such as what themes are emerging. Second, AI plays a part in synthesizing and categorizing the comments for a final report after engagement is closed.

Insights produced through this process depend on the quality of state outreach to bring awareness to the chance to offer input, Marino said, so it was important to have the engagement platform ready for the May 7 announcement. The timing reflects a change from the prior two pilots — both of which had a delay — in an effort to produce higher engagement rates.

The second phase of the latest engagement is slated to begin later this summer. In live forums, a smaller group of Californians representing the state’s workforce will discuss potential policy recommendations in more depth.

A team of ODI researchers, data scientists and analysts has been charged with testing and experimenting to gain these insights, Marino said. Documentation is the other key part of the analysis process; ODI will compile the findings from this project in a public report, which Marino said is slated for a September publication.

INSIGHT-ENABLING TECHNOLOGY


Officials have developed a dashboard to help ODI technologists interact with the data in ways that could produce insights, and ultimately, turn unstructured data into an organized report, Krumrei said. The approach has helped officials find the actionable policy ideas people were raising with their input, even if they were not framed as such.

Notably, the dashboard also allows officials to see if there are gaps in the feedback sources, Marino said, as the goal is to include perspectives from across the state and across industry sectors.

The raw data has also been made available for independent analysis. Open sourcing Engaged California's project data, code and methodology is a “key principle” in this work, Krumrei said.

For other states hoping to tackle engagement through a similar approach, Krumrei recommended working closely with in-house IT teams to validate the processes in alignment with organization-level policy. Collaboration, he said, should include people working directly with such technology tools, including the operations team, designers and communications leaders.

And, he said, user testing should be done in a safe, sandbox environment with test data.

ENHANCING ENGAGEMENT


Other governments may use the Engaged California initiatives as a model, but ODI’s work to improve the state’s public engagement process more broadly may continue, Marino said.

Public comment is legislatively mandated, and compliance is a priority, he said. The goal is to use new tools and technology to ease the administrative burden on state teams and help officials engage more meaningfully with the data they receive.

“This is why ODI was created: we have the flexibility — the agility — to be able to put new tools into practice,” Marino said, noting the ability to change these new solutions if they are not effective.

At the same time, ODI is working to help other departments use new tools and processes to understand and utilize insights from public comment, and to enhance public engagement processes. For example, the ODI team is working with the California Environmental Protection Agency to develop tools that can address its backlog of unexamined public comment.

Engaged California’s methodology may be used elsewhere, but Marino said officials will determine where on a case-by-case basis.
Julia Edinger is a senior staff writer for Government Technology. She has a bachelor's degree in English from the University of Toledo and has since worked in publishing and media. She's currently located in Ohio.