CiviScope — Trustworthy AI for Governance

Public decisions happen across scattered forums, documents, and platforms. People want to participate but can't keep up. AI promises to help—but often obscures more than it reveals.

AI should distribute—not concentrate—power.

Mission

We're building AI that helps people understand and participate in governance. Tools that show their work. Tools that help people decide, not just algorithms.

Because governance you can't understand is governance you can't shape.

Approach

Our AI makes its reasoning visible. You see the evidence it uses. You can verify its logic. You can challenge its analysis.

This means following budget debates, tracing policy changes, understanding who argued what—without a law degree or full-time research staff.

Starting with decentralized organizations, expanding to any institution making public decisions.

Principles

Transparent: The work is shown, not just results
Open: Open source, open methods
Useful: Built for people making decisions
Human-centered: You remain the decision-maker

References

Democratic theory & practice → Schneier & Sanders, Rewiring Democracy (MIT, 2025) ; Landemore, Open Democracy (Princeton, 2020) ; Ostrom, Governing the Commons (Cambridge, 1990)
Participatory systems in practice → vTaiwan / Polis ; Decidim
Provenance & verifiable claims → W3C PROV-O ; W3C Verifiable Credentials 2.0 ; C2PA Content Credentials
Algorithmic accountability & AI governance → Kroll et al., Accountable Algorithms (U. Pa. L. Rev., 2017) ; NIST AI RMF 1.0 (2023)
DAO governance → Buterin, Moving Beyond Coin Voting (2021) ; Ellinger et al., MakerDAO case (SAGE, 2024)
Open government standards → Open Contracting Data Standard ; Open Government Partnership