TLDR: This research paper explores how Generative AI can enhance citizen engagement in participatory budgeting, a process where citizens directly allocate public funds. Focusing on Brazil’s Participe+ platform, the study shows AI’s potential to improve proposal formulation, clarity, and inclusivity by providing structured guidance, data integration, and multilingual support. While highlighting benefits like increased participation and democratic legitimacy, it also addresses critical risks such as algorithmic bias and digital exclusion, advocating for ethical guidelines, hybrid approaches, and continuous monitoring to ensure equitable implementation.
In an era where democratic engagement faces challenges like declining civic participation and increasing societal polarization, a new research paper explores how Generative Artificial Intelligence (AI) can revitalize citizen involvement, particularly in the realm of participatory budgeting. This study, focusing on the Brazilian context and the innovative Participe+ platform in São Paulo, delves into the potential of AI to make democratic processes more inclusive and effective.
Participatory budgeting (PB) is a groundbreaking democratic process that empowers citizens to directly decide how a portion of public funds should be spent. Originating in Porto Alegre, Brazil, in 1989, PB has since spread globally, becoming a flagship example of democratic innovation. It allows residents to propose projects, deliberate on priorities, and vote on funding allocations, leading to tangible improvements in public services and fostering greater transparency and accountability.
The advent of digital technologies has further transformed participatory democracy, giving rise to digital governance and e-participation platforms. These online tools help overcome geographical and time barriers, making participation more accessible. São Paulo’s Participe+ platform, launched in 2020 during the COVID-19 pandemic, is a prime example. Built on the open-source Consul platform, Participe+ dramatically increased citizen engagement in the city’s “Citizen Budget” process, seeing a 500% rise in participation in its first year. This platform allows citizens to submit proposals, comment, and vote online, enhancing transparency and reaching previously underrepresented areas.
Generative AI: A New Frontier for Citizen Engagement
The research highlights Generative AI as a powerful catalyst for further democratic innovation. These AI systems, such as large language models, can produce human-like text and offer novel ways to support communication and creativity in governance. The paper specifically examines how AI can enhance the crucial proposal development phase in participatory budgeting, where citizens’ initial ideas are shaped into concrete, actionable projects.
Traditionally, human facilitators assist citizens in refining their proposals, but this approach has limitations in scalability and consistency. Generative AI can bridge this gap by providing personalized, on-demand assistance. Imagine an AI chatbot guiding a citizen through a series of questions: “What problem do you want to solve?”, “Who would benefit?”, “What resources are needed?”, and “What is the expected impact?”. Based on these responses, the AI can generate a structured draft proposal, ensuring all key components are included.
Beyond structuring, AI can act as a writing assistant, improving the clarity and coherence of proposals. For instance, a simple statement like “The street in front of school is dark and dangerous” could be refined by AI into “Install additional lighting on 5th Avenue near Lincoln School to improve safety for students and pedestrians.” The AI can also suggest specific details and even integrate data from open government systems, such as historical accident rates or estimated costs based on previous projects, making proposals more realistic and evidence-based.
Fostering Diversity and Inclusivity
Generative AI can also encourage a wider range of ideas. Instead of citizens only proposing common projects, AI could suggest alternatives based on their interests, exposing them to new possibilities. Furthermore, AI can help refine and validate proposals by flagging issues like exceeding budget limits or falling outside municipal jurisdiction, and even detect and recommend consolidating duplicate ideas.
Crucially, AI assistance can significantly reduce barriers to participation for citizens with limited literacy, digital skills, or confidence. It can translate informal language into formal proposals and offer multilingual support, making participatory budgeting more inclusive and reflective of diverse urban populations. This aligns with the core principles of democratic innovation: expanding access, reducing inequality, and empowering all voices.
Also Read:
- Understanding AI’s Emotional Leap: Privacy, Ethics, and the Future of Affective Computing
- Designing AI Behavior: PolicyPad’s Collaborative Method for LLM Policies
Navigating the Challenges and Risks
While the potential benefits are substantial, the research also acknowledges significant risks. Algorithmic bias and access inequality are major concerns. If not carefully managed, AI tools could inadvertently favor more digitally literate or affluent citizens, undermining the equity goals of participatory budgeting. The existing digital divide, as seen with the Participe+ platform, could be exacerbated by the introduction of AI.
To mitigate these risks, the paper recommends strong equity measures, including targeted training, support for less digitally literate communities, and continuous monitoring of participation and success rates. Hybrid approaches, combining digital tools with offline outreach and access points, are essential. Ethical guidelines, algorithmic transparency, and accountability mechanisms are also crucial for responsible AI deployment. Governments should consider updating civic participation regulations to formally legitimize AI-assisted input.
Pilot programs and international collaboration, similar to the spread of open-source platforms like Consul, are suggested to gather empirical data and share best practices. The ultimate goal is to ensure that AI supplements, rather than replaces, human interaction, reinforcing the democratic values of inclusion, deliberation, and empowerment.
In conclusion, the integration of Generative AI into participatory budgeting, as explored in this research, represents a meaningful democratic innovation. It builds upon the legacy of PB, reconstructs participatory practices for the digital age, and has the potential to transform how citizens and governments interact in policymaking. For more details, you can read the full research paper here.
