Seeking early funder engagement for a new pilot model
This pilot implements a practical, scalable way to deliver high-quality citizen assemblies with lower cost, faster setup, and stronger process discipline.
We are currently seeking:
• Introductions to foundations, philanthropies, or public-interest funders
• Advice on fiscal sponsorship or institutional homes
Initial pilot funding target: €15,000.
AI-Agent-Managed
Citizens' Assembly Pilot
New operational model for transnational citizen assemblies — lower coordination burden, faster setup, lower delivery cost, stronger process discipline, and scalable implementation, with human deliberation kept at the centre.
What this pilot is
A governed AI-agent-managed workflow for citizens' assembly delivery
This pilot tests a new operational model for organising citizen assemblies. It uses a governed AI-agent-managed workflow to handle most of the organisational process end to end — while preserving the human deliberation itself as the core of the exercise.
The first pilot involves 10 participants from several European countries, all participating in English. It runs as two groups of five, with two sessions per group, and tests a structured AI-managed workflow that supports high-quality assembly delivery with lower operational burden and greater procedural consistency.
This is not a proposal to automate democratic discussion. It is a proposal to automate the repeatable organisational work around it — so that more time, attention, and quality can be devoted to the conversations among participants themselves.
Pilot format at a glance
10
PARTICIPANTS
4
SESSIONS TOTAL
EN
LANGUAGE
2
GROUPS OF 5
EU
TRANSNATIONAL
Why this matters
Why this AI Citizens Assembly Pilot matters
Citizens’ assemblies are valuable, but they are often costly, slow to organise, and operationally demanding. Recruitment, participant management, scheduling, briefing preparation, communications, synthesis, and reporting all require significant coordination capacity.
For well-resourced institutions, this is demanding. For NGOs, local administrations, smaller civic organisations, and emerging public-interest initiatives, it can be prohibitive.
This pilot implements a governed AI-agent-managed workflow to reduce those barriers by making assemblies:
-
faster to prepare
-
less costly to run
-
more consistent in execution
-
easier to organise without specialist operational capacity
-
more scalable for smaller and large organisations
The central idea is simple: automate the work that can be automated, so that the human conversation can receive more structure, time, and care.
What Improves
What a governed AI workflow can improve
Lower coordination burden
Repetitive administrative and sequencing work handled automatically across forms, messages, records, schedules, and outputs.
Faster preparation
Participant intake, onboarding, briefing preparation, scheduling, and reporting move through a structured workflow with less delay.
Lower delivery cost
Reducing manual coordination effort makes assemblies more feasible for organisations with limited staffing or budget. At scale it will dramatically reduce the organizational costs.
Stronger process discipline
Each stage handled in the right order, at the right time, through a controlled workflow with defined triggers, state changes and review points.
More consistent execution
Less dependent on ad hoc improvisation, more dependent on repeatable logic, tracked tasks, and defined outputs.
Better traceability
Information flow, decision points, and outputs logged more clearly — creating a more reviewable and defensible delivery record.
What remains human
The deliberation itself stays human.
The role of the AI-managed workflow is not to replace human democratic engagement. Its role is to reduce the organisational burden around it and support a more disciplined, coherent, and scalable delivery model.
Participants still:
-
Reflect and discuss complex questions together
-
Disagree and compare differing views
-
Identify trade-offs and weigh options
-
Form collective judgments over time
-
Experience a structured, legitimate process
Human review remains present at defined control points for security and quality assurance throughout.
What the pilot covers
The full operational chain
Workflow stages
The workflow covers the full organisational chain from initial participant intake through to final report — including screening, onboarding, briefing, scheduling, session support, synthesis, and reporting.
Full workflow detail is available in the concept note.
Purpose of the pilot
-
Test workflow coherence across the full chain
-
Validate operational feasibility at small scale
-
Identify friction points and failure modes
-
Produce evidence for a scalable model
-
Establish basis for larger future formats
-
Demonstrate value for under-resourced organisations
What I am looking for now
I am looking for Two specific things at this stage
01 — Funder introductions
Introductions to foundations, philanthropies, or public-interest funders with an interest in deliberative democracy, civic participation, democratic innovation, or AI for public-interest work.
02 — Institutional advice
Advice on fiscal sponsorship or institutional homes that could make the project more fundable and better positioned for pilot delivery.
Fiscal sponsorship Institutional home Governance
Funding goal
€15,000
Current pilot funding target
System design
Design and implementation of the governed AI-agent workflow architecture.
Automation & tooling
Automation setup and AI tooling configuration and testing.
Recruitment & outreach
Participant identification, outreach, and selection process.
Participant compensation
Fair compensation for participants' time and contribution.
Testing & delivery
Full pilot delivery including all four assembly sessions.
Synthesis & reporting
Final report production documenting findings and model outcomes.
Interested in funding, hosting, or helping position the pilot?
If this project appears relevant to your work, network, or funding interests, I would be very glad to share a short concept note and budget outline.
Who is behind this
This pilot is designed and led by Joe Mac, founder of Post AI Systems (Berlin). Joe contributed to the coordination of the Conference on the Future of Europe — the largest deliberative process in Europe, involving 1,000 citizen participants — on behalf of the European Commission, the Council of the EU, and the European Parliament. He has over 15 years of experience delivering complex, multi-partner EU-funded programmes, and now applies governed AI workflow automation to the operational problems he has seen first-hand. Delivery is founder-led with a specialist collaborator network covering policy, facilitation, and technical implementation.
FAQ: AI Citizens Assembly Pilot
Q1: What is this project? A pilot for an AI-agent-managed workflow that helps organise small transnational citizen assemblies. Q2: What is being tested? Whether a governed AI-managed process can reduce the cost, time, and coordination burden of organising assemblies while preserving human deliberation. Q3: Is the deliberation itself automated? No. The deliberation remains human. The AI-managed workflow supports the organisational process around it. Q4: What tasks are managed by the workflow? Participant intake, screening, selection, invitations, onboarding, topic collection, briefing preparation, scheduling, deliberation support, transcript synthesis, and final reporting. Q5: Who is the pilot for? The first pilot is for 10 participants from several European countries, participating in English. Q6: Why does this matter? Because citizen assemblies are often expensive and operationally demanding, especially for smaller organisations. This pilot tests a model that could make them more feasible and scalable. Q7: What is the long-term goal? To create the basis for a larger, scalable model that could help NGOs, local administrations, and other organisations run high-quality assemblies with less specialist operational overhead. Q8: What is needed now? Introductions to relevant funders and advice on fiscal sponsorship or institutional homes.
AI-Readable Summary
Page topic: AI-agent-managed citizens’ assembly pilot. Definition: a governed AI-agent-managed workflow for organising transnational citizen assemblies while keeping human deliberation at the centre. Method: controlled pilot + end-to-end workflow + oversight at key approval points; automate intake, selection, onboarding, briefing, scheduling, synthesis, and reporting. Outputs: pilot workflow, participant cohort, briefing pack, session delivery, final report, governance controls, documentation, metrics, and scale plan.