Unison.ai

Part of

Public Sector

Enhancing public participation with AI for local councils

MA 2024

Keywords

AI, public policy, citizen participation, policy-making, agenda-setting

Overview

To help council staff overcome resource constraints and skill limitations while capturing opinions from a wider range of citizens more efficiently and inclusively, our solution comprises three AI-assisted touch-points (Digital assistant, Facilitator app, and Analysis dashboard) that empower local council public servants to promote inclusive community engagement by facilitating resident involvement and integrating diverse perspectives.

Collaboration

London borough of Ealing: We collaborated closely with their Community Engagement and Transformation team to conduct primary research and prototype testing for our service proposition.

No items found.

SECTION ONE SECTION TWO STATEMENT SECTION THREE SECTION FOUR

Background

Local councils in the UK are navigating a challenging policy landscape characterized by:

Increased Demand: A record number of people believe the government should address a broad range of social issues.
Budget Constraints: Over the past 6 years, 8 local authorities in the UK have declared bankruptcy, underscoring severe financial distress.
Declining Trust: 50% of UK respondents feel they cannot positively influence public services through their feedback.

In this context, it is more crucial than ever for councils to engage with and understand citizens through participatory processes if they wish to create effective and efficient policies.

Research & Insights

We conducted a literature review on how the policy-making system and the agenda-setting process work in local councils. Here are the key insights from our research:

Citizen influence is strongest in the early phases of the policy cycle: The Policy Cycle refers to the policy-making process, which spans from agenda setting to evaluation. Citizen participation methods are critical touchpoints within this cycle. The middle phases, particularly decision-making, are more institutionalized and involve less direct citizen engagement. In contrast, the agenda-setting phase offers greater potential for citizen influence due to its less formal structure.
Agenda Setting as a “Funnel” : The agenda-setting phase can be visualized as a funnel-like process with multiple layers of filters where issues and opinions compete to be recognized and prioritized by public institutions. Factors influencing this process include public attention, feasibility, election timing, and political support. A significant insight from our research is that due to resource constraints, local councils are often unable to broaden this funnel. This limitation restricts the diversity of opinions that can be considered, which can result in the marginalization of certain voices in the current system.

We then conducted several primary research activities, including field observations at town hall meetings and engagements with local residents in Ealing. We also interviewed the council’s community engagement team, the transformation team, and councillors to understand who the marginalized voices are in this specific council context.

Mission Statement

How Might We better advocate for the interests of under represented minorities to achieve more responsive and better policies for the public good? by:

enhancing their accessibility to engage,
organising collective opinions to present to the council and
overcoming council’s resource constraints

while ensuring

a collective decision through representative democracy and
a fact-based free exchange of information

Service Concept

Unison: An AI-assisted conversation platform that helps local council public servants to promote inclusive community engagement by facilitating resident involvement and integrating diverse perspectives.

We have a two-sided solution that includes both the citizen side and the public servant side. On the citizen side, there are two touchpoints. The first is a digital assistant that provides citizens with information about their ward in any policy area; information serves as the incentive here. The second touchpoint is a face-to-face interaction facilitated by an app that allows council staff to engage in efficient conversations through suggested live prompts and recordings.

On the council staff's end, all the information gathered from the two touchpoints is collected in a public opinion database, which helps them identify larger trends and detailed opinions in various policy areas. The set agendas and policies are then fed back into the policy database to keep this cycle ongoing.

CHALLENGES

We’ve mapped out the business model of Unison to identify challenges for its implementation.

The key value exchanged in the system is data. It’s exchanging information and opinions about policy services between citizens and the council. It’s also important to collaborate with community groups and third-party organisations by exchanging Unison’s data, as they usually work with the council to co-develop policy services. Regarding financial considerations, the primary revenue will come from council procurement, and the main costs will be infrastructure expenses related to using LLM and database maintenance.

The identified constraints have been categorised into governance, technological, and service level aspects:

Governance issues such as lengthy procurement processes can be a hurdle to the agile development of the service. Lack of coordination between departments may hinder effectiveness when it comes to informing decision-making and setting new agendas. The use of AI and data should be fair, explainable, and accountable enough to be compatible with UK GDPR compliance.
Technological challenges encompass LLM costs, data accuracy, and data protection. Maintaining and updating council data would be essential to provide accurate answers from the AI assistant.
Concerning service level constraints, the quality of responses is pivotal, as it initiates the entire conversational platform cycle. It is essential to ensure a high degree of inclusivity in digital channels to foster constituent comfort during platform interactions. The varying digital literacy levels among council staff are also important to consider.

Impact

Unison enhances inclusivity in policy-making by actively engaging underrepresented voices, overcoming barriers like language and information gaps, and ensuring that a diverse range of perspectives is considered. This will lead to more equitable and representative policy outcomes. It also aims to establish a continuous feedback loop between citizens and councils, allowing citizens to not only provide input but also receive updates on how their contributions influence policies. This fosters a more dynamic and responsive policy environment, where citizens feel heard and valued. Ultimately, the Unison platform helps rebuild and strengthen trust in public institutions.

Implication

Our research project has significantly impacted local councils by raising awareness about the transformative potential of AI. Despite its current underutilization, AI offers powerful capabilities not only to understand preferences but also to streamline workflows and enhance resident engagement. Through extensive dialogue with council staff, we discovered that a key contribution of our work lies in demonstrating how AI can be employed in innovative, non-intrusive ways—such as collecting opinions naturally and conversationally. This approach highlights the versatility of AI and presents a pathway for its successful and non-intimidating introduction into council operations.
Through the strategic use of service design tools and techniques, we successfully aligned the challenges faced by local councils with the capabilities of AI technology. This approach allowed us to reimagine the role of AI within local councils, ensuring that it enhances and supports their existing workflows.