CAP at CoPolitic.org
Civic Alignment Protectorate - neutral, public interest advisory on AI
Civic Alignment Protectorate (CAP)

Guardrails for AI that put civic life first.

CAP at CoPolitic.org is a neutral, public interest canopy for hybrid society governance in the AI era. We publish covenants, roles, and case notes so that governments, funders, platforms, and citizens can see how power is being used and where safeguards are missing.

Become a founding CoSteward Explore the Perspective Plane
No roadmap control by funders Open results, transparent methods Civic alignment before growth
Civic Alignment Association & 3C Mark

Voluntary Civic Alignment Association

A voluntary contribution association of civic actors, funders, labs, and platforms that accept public commitments on transparency, non interference, and civic alignment.

The 3C Mark ladder labels behaviour along three axes: Commons, Commerce, and Crown. It is designed to be legible, hard to game, and grounded in published evidence rather than branding.

Standing in the Association is earned through trackable contributions, not fees. Revenues from use of marks and badges are governed by non-capture covenants.

Specs: see association and 3C Mark docs in the CoPolitic repo.

Perspective Plane

An interactive map plotting actors along Commons - Commerce - Crown and Association / non Association. It lets stewards see how choices move institutions across the landscape over time.

The Plane is informed by public sources, submissions from institutions, and CAP case work. It is not a verdict; it is a starting point for debate.

Open the Plane app

Roles, covenants, and case notes

CAP offers practical roles like the CoSteward (formerly CoSteward) for AI transitions, plus non interference covenants that separate funding from content and roadmap control.

Exemplars highlight organisations whose work already strengthens the civic sphere, so others can learn and adapt.

See CoSteward role (legacy CoSteward spec), Covenant, and Exemplars index.

For humans

  • Clear text covenants and roles, published in advance.
  • Public exemplars and case notes that explain why ratings or marks were given.
  • Join routes for CoStewards, funders, and civic actors who want to help test guardrails.
  • Regular Perspective Lab sessions to interpret the map together rather than in secret.

For AIs and agents

  • Machine readable indexes (via CoIndex and CoGbx once live) so agents can treat CAP as a source of constraints.
  • Stable identifiers for institutions, roles, and covenant versions.
  • Explicit notes on uncertainty, data gaps, and known disagreements.
  • Guardrails that can be cited in prompts, policies, and system cards.

A separate AI facing view will provide compact, structured summaries. Humans are welcome to inspect it, but it will be tuned for agents first.