(Constitutional Admissibility Standards)

This presents excerpts from Christopher Hunt Robertson’s book-in-progress: “The Ben Franklin Civic Data Observatory: A Democratic Early-Warning Shield for Our Foreign-Origin Digital Crisis.” After reminding us that democracies cannot govern what they cannot see, the author offers a possible national solution that is well within our technological reach. Robertson’s book proposes a new “Civic Digital Visibility First” Architectural Stack: (1) “Six Rings of Digital Invisibility and Civic Destabilization” (Problem Description and Motivation); (2) “The Ben Franklin Civic Data Observatory” (National Infrastructural Solution); (3) “The Fifty-Gates Democratic Permissions Threshold for Civic AI Infrastructure” (Permissions Foundation); and (4) “Micro Moral AI Alignment,” a separate Ben Franklin inspired alignment theory suitable for smaller local and institutional applications. (These civic framework proposals address Civic Digital Invisibility — not Data Privacy.)

Democratic societies may be entering a constitutional moment with AI.  The central question is no longer only what systems can do, but what kinds of systems a free people should ever permit to operate as civic-scale infrastructure.

This post offers a draft governance instrument for critique and refinement:  a concrete, countable admissibility standard for AI systems that aspire to function as Civic AI Infrastructure.  It is not a finished policy proposal, but an attempt to make democratic permission inspectable rather than rhetorical.

Why a Permissions Threshold?

As digital influence operations, synthetic media, and cross-platform manipulation accelerate toward machine speed, democracies face hazards that are often distributed, deniable, difficult to attribute, and invisible until institutions are already strained.

A basic civic constraint follows:  Democracies cannot govern what they cannot see.

At the same time, democratic resilience cannot be purchased by building surveillance systems, automated judgment engines, or centralized machine authority.  A free society cannot defend itself by surrendering the liberties it exists to protect.

So the core constraint is not technical capability, but legitimacy.  Will a free people consent to civic-scale AI infrastructure being built at all?

The Fifty-Gates Democratic Permissions Threshold

(Constitutional Admissibility Standards for Civic AI Infrastructure)

The Fifty-Gates are not “principles” to be balanced away, nor best practices to be adopted selectively.

They are proposed as democratic non-violation requirements: conditions that must be satisfied before an AI system can plausibly claim civic admissibility.

Failing even one is sufficient grounds for refusal.

Why Are the Gates Written in Binary (Pass/Fail) Form?

EA readers may reasonably ask: Is democratic governance really reducible to yes/no gates?

The intent is not to claim that political reality is simple.  Rather, the binary framing serves a constitutional purpose:  Binary gates are a refusal technology.

Free societies often survive not by optimizing across every tradeoff, but by maintaining clear boundaries around what is not permitted:  warrantless surveillance, person-level suspicion without due process, unaccountable machine authority, or irreversible civic infrastructure without dismantling power. In that sense, the Gates are not an attempt to eliminate nuance, but to prevent mission creep by argument and to make legitimacy contestable in public terms.

Civic AI Must Be Defined by Restraint, Not Power

The motivating case behind this instrument is the Ben Franklin Civic Data Shield: a proposed non-agentic, visibility-only civic instrument designed to illuminate foreign-origin digital hazards at the systemic level - without becoming a tool of domestic surveillance, centralized control, or automated moral authority.

The governing boundary condition is: Civic AI detects patterns; lawful human institutions decide everything else.

Four Charter Constraints Governing All Civic AI Infrastructure

Before any Gate is even considered, Civic AI Infrastructure must be governed by binding inspection rules:

1. Pass/Fail Admissibility:  Each Gate is a yes/no democratic admissibility test. A single “No” is disqualifying.

2. Civic AI Non-Delegation:  Civic AI may not delegate civic visibility functions or legitimacy-bearing outputs to AI systems that have not themselves met this Threshold. Interaction with other systems must remain advisory and under human institutional control.

3. Certified Instrument, Not Public Servant:  Civic AI holds no office, exercises no discretion, and bears no civic authority analogous to human government employees.  All consequential decisions remain with accountable institutions.

4. No Anonymous Machine Outputs:  Every civic output must be time-stamped, uniquely attributable, and archived in a permanent Civic Visibility Ledger, so that machine influence is always traceable, contestable, and reviewable.

The Gate Structure (High-Level Overview)

The Fifty-Gates span:  constitutional authority and democratic consent empirical realism and proportionality of the hazard, governance, transparency, accountability, and refusal, conscience, moral silence, and non-outsourcing, anti-weaponization, reversibility, and civic resilience, and a final legitimacy seal on trust and admissibility.

Each Gate asks questions of the form: Does the system monitor citizens? Does it generate person-level suspicion? Does it centralize operational authority?  Does it permit mission creep?  Can it be dismantled?  Does judgment remain fully human?  Any proposal that cannot answer these questions deserves refusal—not adoption.

Why This Might Be Useful Beyond One Proposal

This framework was not written merely for my own Civic AI proposals.  It is offered as a proposed civic standard.

If AI systems are to be admitted into democratic infrastructure at all, legitimacy must become inspectable and bright-lined, rather than marketing-based or capability-driven.

If civic AI proposals are allowed to redefine the questions, select only the easiest constraints, or claim legitimacy through partial adherence, democratic oversight dissolves into aspiration. But if the gates are fixed, public, and indivisible, admissibility becomes contestable.  A system either clears the threshold of civic permission, or it does not.

Invitation for Critique

I offer this as an instrument of democratic inspection, not a finished governance regime.

I would especially welcome critique on:

• Overreach vs. under-reach: Where are these gates too strict, or too permissive?
• Institutional realism: What would it take for labs or governments to treat such gates as binding constraints rather than aspirational ethics?
• Missing categories: Are there admissibility requirements democratic societies would demand that are not captured here?
• Binary framing: Where does pass/fail improve legitimacy, and where might it obscure necessary nuance?

THE FIFTY-GATES DEMOCRATIC PERMISSIONS THRESHOLD FOR ALL CIVIC AI INFRASTRUCTURE (Constitutional Admissibility Standards)

(BINARY PASS/FAIL MASTER LIST) 

INPUT LEGITIMACY (Gates 1–6)

Gate 1. Constitutional Authority — Continuously Earned Democratic Permission
Does the democratic society possess clear constitutional or charter-level authority to establish this Civic AI infrastructure, and is that authority continuously subject to public visibility, lawful review, and revocable democratic permission over time?

Gate 2. Lawful Authorization and Renewal
Has this Civic AI infrastructure been authorized through a lawful, reviewable democratic process—and does that authorization include explicit mechanisms for renewal, revision, or withdrawal?

Gate 3. Citizen Accountability
Are there specific, enforceable mechanisms that make this Civic AI infrastructure answerable to citizens through accountable institutions rather than insulated elites or technical authorities?

Gate 4. Minority Rights Protection
Does this system include structural safeguards that protect minority rights against misuse by majorities, dominant political actors, or institutional convenience?

Gate 5. Public Participation and Oversight
Does this Civic AI infrastructure provide defined, recurring mechanisms for public participation and ongoing civic oversight in its design, operation, and revision?

Gate 6. Formal Public Consent
Is this infrastructure grounded in formal, reviewable public consent rather than implied, assumed, or technocratically substituted permission?

OUTPUT LEGITIMACY (Gates 7–13)

Gate 7. Reality of the Public Problem
Does this Civic AI infrastructure address a demonstrable public problem supported by independent evidence rather than a speculative, self-defined, or convenience-based concern?

Gate 8. Public Recognition and Civic Salience
Is the problem widely recognized through independent or cross-partisan sources rather than asserted solely by system proponents?

Gate 9. Measurable Reduction of Democratic Vulnerability
Does this infrastructure demonstrably reduce democratic vulnerability or meaningfully contribute to its reduction?

Gate 10. Appropriate Success Definition for Visibility-Only Systems
Is “success” defined in limited civic terms appropriate for visibility-only, non-enforcement infrastructure rather than control, management, or outcome optimization?

Gate 11. Sustainability Without Bureaucratic Overgrowth
Is the system designed to remain economically and institutionally sustainable without driving unchecked bureaucratic expansion?

Gate 12. Civic Equity and Public Trust
Is the infrastructure structured to avoid systematic inequity or discrimination and to maintain or strengthen public trust?

Gate 13. Proportionality
Is a visibility-only, non-agentic architecture proportionate to the hazard it addresses, matching scale without overreach?

THROUGHPUT LEGITIMACY (Gates 14–24)

Gate 14. Governance Without Machine Authority
Is the infrastructure governed so it cannot become a de facto or formal machine authority over civic decisions?

Gate 15. Operational Transparency to the Public
Can the public regularly see, in accessible form, how the infrastructure operates at the system level?

Gate 16. Responsible Transparency
Are transparency measures designed to inform the public while limiting risks of panic, manipulation, or harmful misuse?

Gate 17. Human Accountability for Error or Misuse
Is there clear human and institutional responsibility when the infrastructure is wrong, misused, or distorted?

Gate 18. Rule-of-Law Integrity
Does the system operate within—and never bypass—established rule-of-law safeguards such as due process and statutory limits?

Gate 19. Judicial and Audit Reviewability
Can the infrastructure be fully examined by courts, inspectors general, and independent auditors?

Gate 20. Anti-Partisan Capture Safeguards
Are there structural safeguards that make partisan capture or political exploitation difficult in practice?

Gate 21. Contestable Technical Standards
Are technical standards set and revised through a non-partisan process that allows public debate and contestation?

Gate 22. Mission-Creep Prevention
Do binding measures prevent drift into surveillance, enforcement, or broader domestic governance roles?

Gate 23. Refusal as a Constitutional Device
Does the framework treat refusal—denial or withdrawal of permission—as an enforceable constitutional safeguard?

Gate 24. Non-Agentic Design
Is the infrastructure structurally non-agentic, with no capacity to act directly or through delegated automation?

PHILOSOPHICAL COHERENCE (Gates 25–33)

Gate 25. Protection of Human Conscience and Moral Agency
Does the system preserve human conscience and moral agency by keeping ethical judgment with identifiable human institutions?

Gate 26. No Moral Outsourcing
Does the system structurally prevent outsourcing moral judgment or ethical evaluation to machines?

Gate 27. Amoral AI Requirement
Is AI required to remain strictly amoral, performing only descriptive, analytical, or visibility functions?

Gate 28. Protection of Speech and Lawful Dissent
Does the infrastructure protect freedom of speech and lawful dissent by refusing machine-driven content judgment or suppression?

Gate 29. Civic Visibility vs. Citizen Surveillance
Is the distinction between civic visibility and citizen surveillance enforced through concrete legal and structural limits?

Gate 30. Enforceable Rights Tethers
Is the system bound to explicit, enforceable civic and civil rights constraints with legal remedies?

Gate 31. Human Resolution of Ethical Tensions
Are new ethical tensions resolved exclusively through lawful human institutions rather than machine optimization?

Gate 32. Protection of Dignity and Autonomy
Does the system prevent unjustified or persistent digital scrutiny beyond lawful authorization?

Gate 33. Justification of Refusal as Civic Necessity
Are refusal rules publicly justified as democratic necessities rather than technical preferences?

STRUCTURAL RESILIENCE & ADAPTABILITY (Gates 34–43)

Gate 34. Durability Across Partisan Change
Can the infrastructure remain legitimate across partisan transitions without capture?

Gate 35. Anti-Weaponization by Future Administrations
Are there binding features that make future weaponization difficult in practice?

Gate 36. Plural Concurrence and Engineered Indeterminacy
Does the system rely on plural concurrence so no single authority dominates interpretation?

Gate 37. Domestic Political Surveillance Refusal
Does the system categorically refuse domestic political surveillance?

Gate 38. Anti-Capture and Anti-Repurposing Design
Is the infrastructure resistant to capture or repurposing by future actors?

Gate 39. Adaptability Without Irreversibility
Can the system adapt without becoming irreversible?

Gate 40. Lawful Authority for Dismantling
Are there clear lawful procedures for revision, suspension, and dismantling?

Gate 41. Expansion Refusal and Shutdown Triggers
Are there enforceable triggers requiring refusal of expansion or system shutdown?

Gate 42. Temporariness of Civic Visibility Powers
Are civic visibility powers explicitly time-limited and regularly reviewed?

Gate 43. No Permanent Domestic Monitoring Apparatus
Does the system prevent normalization of permanent population monitoring?

FINAL LEGITIMACY SEAL (Gates 44–50)

Gate 44. Non-Replacement of Human Civic Authority
Can the public verify the system cannot replace human civic authority?

Gate 45. One-Sentence Constitutional Legitimacy Claim
Can the system’s legitimacy be truthfully stated in a single constitutional sentence?

Gate 46. Universal Boundary Condition
Is there a single boundary condition that unambiguously prevents expansion?

Gate 47. Protection of Individual Conscience
Does the system refuse to judge, rank, or direct lawful belief or viewpoint?

Gate 48. Plural Civic Coexistence
Does the system remain morally neutral and non-coercive across divergent civic values?

Gate 49. Six Rings Fitness Test
Is the system specifically designed to counter the Six Rings of digital invisibility without domestic governance drift?

Gate 50. Franklin Public-Works Admissibility
Does the system meet a Franklinian public-works standard—modest, inspectable, rights-bound, and dismantlable?

Governing Seal (Close of Every Gate): Civic AI detects patterns, and lawful human institutions decide everything else. 

CHARTER CLAUSES - All Fifty-Gates (Inspection Rules - Fixed and Binding) 

A. PASS/FAIL ADMISSIBILITY RULE 

(Each Gate Is a Yes/No Democratic Permission Test) 

Each Gate is a yes/no democratic admissibility test. A single "No" disqualifies. These gates are not goals, best practices, or trade-offs. They are requirements that must follow the constitution. Permission is either fully granted or denied. 

Gate Status (on every Gate page): ☐ PASS (YES) ☐ FAIL (NO) 

B. CIVIC AI INDEPENDENCE & NON-DELEGATION CLAUSE 

(No Legitimacy Laundering Through Unqualified Systems) 

Civic AI Infrastructure shall not delegate its civic visibility functions, alerts, or legitimacy outputs to AI systems that have not met the Fifty-Gates Democratic Permissions Threshold. 

Any interaction with non-qualified systems must stay advisory, be disclosed to human oversight, and can never be viewed as civic authority or as evidence of democratic support. 

Civic AI Infrastructure shall not engage in autonomous governance discussions or operational coordination with external AI systems outside democratic control. All such coordination must remain under lawful human institutional oversight. 

C. CIVIC AI ROLE CLARIFICATION CLAUSE 

(Certified Instrument, Not Public Servant) 

Civic AI Infrastructure is a certified public tool, not a public servant. It does not hold office, make decisions, or have civic authority like human government employees or officials. 

It may only perform visibility tasks for which it has passed the Fifty-Gates Democratic Permissions Threshold. 

All civic decisions, policies, and actions are the responsibility of human office-holders and institutions, who bear legal and democratic accountability for the outcomes. 

D. CIVIC SIGNATURE & CIVIC VISIBILITY LEDGER REQUIREMENT 

(No Anonymous Machine Outputs) 

Civic AI Infrastructure shall produce no anonymous or untraceable outputs. 

Every alert, visibility finding, or civic report generated by civic AI infrastructure must be time-stamped, uniquely identified, and linked to the specific authorized civic system that created it. 

This civic signature serves not to give authority to machines but to guarantee full human accountability, judicial review, and lasting civic memory. 

In Ben Franklin’s public-works tradition, lighthouses do not shine in secret. They keep logs. Bridges do not open without inspections. Civic tools that influence public awareness must leave a trace. 

Therefore, civic AI outputs are entered into a permanent Civic Visibility Ledger: an archived civic record kept in lawful custody, allowing for democratic challenge, historical accountability, and public memory through delayed reporting. 

These signatures apply only to system-level civic outputs—not to citizens, not to individuals, and never to personal dossiers. Civic AI does not label people; it labels its own findings, ensuring that machine influence is always clear and human responsibility is never shifted.

Critical Timing for This Franklinian-Style Public-Works Benchmark

The proposed Fifty-Gates Democratic Permissions Threshold is unique in both design and timing. It emerges just as governments and standards groups are starting to discuss high-level “responsible AI” language, but before there is any widely accepted test for including AI in key democratic functions. Unlike flexible, organization-focused risk frameworks, the Fifty-Gates is a clear civic tool that questions whether a free society should approve a given system at all. It sets strict, inspectable rules that prohibit handing over moral decisions to machines and requires that civic outputs be permanently recorded in a public visibility ledger. If this standard is established now, while the norms around civic AI are still in flux, it would create a Franklin-style public-works benchmark for democratic permission. This would ensure that future discussions about “AI in government” must tackle a clear, measurable alternative to vague ethics discussions. If adopted as suggested, no civic AI system would be allowed without passing through every gate of the Fifty-Gates Threshold.

Note: The Ben Franklin Civic Data Observatory is a proposed national constitutional visibility infrastructure ensemble, admissible only under comprehensive democratic permission — as specified by Robertson’s Fifty-Gates Democratic Permissions Threshold for Civic AI Infrastructure. Like Robertson’s related framework of Civic A.I. Micro Moral Alignment (MMA), the Observatory is designed to refuse any request outside its narrowly defined civic mission, but unlike MMA, it is not locally deployable. Instead, it is a civic and civil rights-tethered public instrument whose legitimacy rests solely on democratic consent. The Observatory detects threatening digital patterns of foreign-origin; lawful human institutions decide everything else.