AI x Democratic Societies

Context

Artificial intelligence is increasingly reshaping democratic societies. While AI creates new opportunities for civic participation, access to information, and institutional transparency, it is also accelerating misinformation, concentrating power, enabling surveillance, and challenging public trust.

Democracy depends on trustworthy information, accountable institutions, and the ability of people to participate meaningfully in civic life. As AI and the vast array of data-driven technologies become ever-more embedded in our day-to-day, the question is no longer whether AI will shape democracy, but how AI systems can be built to support democratic values by design.

Mozilla Foundation has long advocated for an open, trustworthy internet and is investing in technologies that strengthen democratic practice through its Democracy × AI Incubator. Many promising technology projects struggle to move beyond early development towards real-world implementation and institutional adoption; that is what Mozilla Foundation’s incubator supports: they bridge the critical gap between promising prototype and sustainable project.

The Prototype Fund Switzerland’s edition on “Responsible and Sustainable AI” provides a complementary and Swiss-based environment to Mozilla Foundation’s global incubator activity: AI & Democratic Societies ideas will be translated into prototypes, alongside and feeding into governance approaches and concrete societal implications.

Rather than viewing Responsible AI as an abstract policy objective, this challenge explores – based on practical insights and experimentation – how we can further foment democratic institutions, e.g., through local or community-owned data approaches or AI models, public-interest technology, and civic participation.

Current AI ecosystems, driven by commercial optimization and platform incentives, need to reorient their attention and build in desired democratic outcomes by design.

So far, AI- and data-driven technologies have severely impacted democracies globally:

• Societies globally – from citizens to senior decision-makers – struggle to distinguish trustworthy from synthetic or manipulated information;
• Democratic participation becomes yet again fragmented by opaque algorithmic systems;
• Institutions lack practical tools to responsibly integrate AI, from SMEs to large-scale organizations;
• Communities continue to have limited influence over how AI systems are designed and deployed;
• Promising civic technologies frequently fail to bridge the gap between prototype and long-term adoption.

There is a need to prototype AI systems that actively strengthen weakened democratic institutions or social pacts: This cannot be done by merely promoting democratic values or narratives – but by building and enabling AI- and data-building practices that aim at building ever more valuable every-day-lives for today and tomorrow – despite being increasingly exposed to digital technologies by default.

Challenge Framing

This challenge invites interdisciplinary teams to prototype AI-enabled technologies that strengthen democratic participation, transparency, institutional accountability, and public-interest technology.

Projects should explore how AI can support democratic societies through practical, scalable, and community-centered solutions.

Potential directions include:

Better Information

• trustworthy information ecosystems
• collective fact verification
• civic information assistants
• multilingual trustworthy summaries
• algorithmic transparency tools

Transparent & Accountable Institutions

• public-sector transparency tools
• legislative and policy tracking
• public data interfaces
• participatory decision-support
• deliberation and consensus-building platforms

Civic Space & Democratic Participation

• privacy-preserving civic technologies
• community governance platforms
• tools supporting civic participation
• AI supporting local democratic processes
• technologies protecting civic spaces

Democratic AI Infrastructure

• locally governed AI models
• community-owned AI infrastructures
• public-interest AI services
• governance mechanisms for responsible deployment
• sustainable funding models for democratic technology

Projects are encouraged to combine technical innovation with governance experimentation, ensuring that responsible deployment is considered from the outset.

Key Concepts / Lens

• Responsible AI
• Democratic and/or local AI
• Public Interest Technology
• Civic Technology
• Transparency & Accountability
• Community Governance
• AI & Democracy
• Digital Rights
• Open Innovation

Expected Outputs

Projects are expected to develop:

• a functional technical or socio-technical prototype;
• a clearly defined democratic or civic use case;
• documentation of assumptions, governance considerations, and implementation learnings;
• a pilot concept demonstrating how the solution could be deployed in practice;
• recommendations for responsible implementation within democratic institutions or communities.

Expected Outcomes

Tier 1 – Prototype Impact

• Working AI prototypes supporting democratic participation or institutional transparency
• Evidence for responsible deployment of AI in democratic contexts
• Practical governance models accompanying technical solutions
• Increased understanding of democratic AI infrastructures

Tier 2 – Systemic Impact

• Policy-relevant insights for democratic AI governance
• Contributions to international discussions on AI and democracy
• New approaches connecting civic technology, AI governance, and public-interest innovation
• Demonstration that democratic alternatives to mainstream AI ecosystems can be technically viable and socially valuable

Key Questions

• How can AI actively strengthen democratic practice?
• What makes AI systems accountable to communities rather than platforms?
• How can AI improve access to trustworthy information?
• Which governance approaches enable community-centered AI deployment?
• How can civic technologies become sustainable beyond the prototype stage?
• What does Democratic AI Infrastructure look like in practice?

---

Challenge Owner

Mozilla Foundation