FOUNDATION Series · White Paper
Congressus: A Framework for AI-Mediated Democratic Governance
For Any Entity. At Any Scale. Across Any Era.
Version
2.0 — Pre-Draft Latest
Published
May 2026
Status
Open for peer review & collaboration
Contents
Abstract
Democratic governance systems worldwide face a structural crisis. Institutions designed for a slower, more localised world are failing to address the speed and scale of contemporary challenges. This paper presents Congressus: a complete architecture for AI-mediated democratic governance that does not replace human decision-making but restructures when and how humans engage with decisions that affect them.
Version 2.0 adds three significant contributions: a philosophical foundations section establishing Congressus as the successor to religion’s historical role as civilisation’s moral coordination mechanism; the Precision Principle, a formal drafting standard requiring legally defined terms in all policy proposals; and Article VII — the Private Sphere Doctrine, establishing that every citizen retains absolute sovereignty over their private sphere.
1. Introduction
1.1 The Governance Crisis
Democratic governance systems globally are experiencing a structural crisis that transcends partisan politics. The crisis is architectural: institutions designed for a slower, more localised world are systematically failing to address the speed and scale of contemporary challenges.
The symptoms are visible across all democratic systems: legislative paralysis on issues where citizen consensus exists, systematic capture of policy by concentrated interests, inability to optimise for long-term outcomes under short electoral time horizons, and the structural amplification of extreme positions at the expense of majority preferences that typically occupy a more nuanced middle ground.
The abortion debate in the United States illustrates the failure mode precisely. Consistent polling shows approximately 73% of Americans hold nuanced positions — supporting access under some circumstances with some limitations. Yet the political system has produced a binary outcome driven by the approximately 27% at the ideological poles. The majority’s actual position has never been implemented because the system has no mechanism to find it, represent it, or negotiate toward it.
1.2 The Congressus Response
Congressus integrates three converging developments — large language models capable of policy reasoning, mature multi-agent negotiation frameworks, and demonstrated digital democracy experiments — into a complete governance architecture that restructures when and how citizens engage with decisions affecting them.
Version 2.0 adds the philosophical and semantic foundations that complete the architecture: an account of why the Congressus Floor carries genuine moral wisdom without requiring supernatural justification; a precision drafting standard that eliminates the semantic ambiguity through which ideological conflicts are manufactured; and the Private Sphere Doctrine that defines the boundary between legitimate collective governance and impermissible ideological imposition.
1.3 Contributions of This Paper
In addition to the original contributions documented in Version 1.0, Version 2.0 adds:
- A philosophical foundations account connecting Congressus to the history of moral coordination in human civilisation, positioning it as the successor to religion’s coordination role.
- The Precision Principle: a formal drafting standard requiring legally defined terms in all policy proposals.
- Article VII — The Private Sphere Doctrine: a constitutional provision establishing absolute citizen sovereignty over the private sphere, with a two-part harm test.
- The Citizenship Frame: a precise rights architecture using legally defined citizenship rather than philosophically contested personhood.
- A complete edge case analysis of the Private Sphere Doctrine across seven categories of genuine difficulty.
2. Related Work and Positioning
2.1 Deliberative Democracy Theory
The theoretical foundation for citizen deliberation as a source of legitimate governance derives from Habermas’s communicative action theory, Rawls’s concept of overlapping consensus, and Fishkin’s empirical work on deliberative polling. Rawls’s political liberalism is directly relevant to the Private Sphere Doctrine: his concept of public reason — that laws governing citizens must be justifiable in terms accessible to all citizens regardless of their comprehensive moral doctrine — is the philosophical foundation for the Private Sphere Doctrine’s two-part harm test.
2.2 Moral Development Theory
Kohlberg’s stages of moral development provide the empirical foundation for the philosophical foundations section. Kohlberg demonstrated that moral reasoning develops from rule-following under threat of punishment through social conformity to genuine principled reasoning. Congressus is designed to create the systemic conditions under which principled reasoning becomes the dominant mode — not by requiring it but by making it the rational self-interest choice.
2.3 Multi-Agent AI Systems
Zhu et al. (2026) demonstrated that AI delegate agents in multi-party bargaining produce Pareto-improving proposals generating positive externalities even for non-adopting participants. Lavi (2025) proposed the One Person, One Bot model of direct democracy through AI delegation. The Democracy-in-Silico project (2025) validates simulation methodology for AI governance systems. Congressus extends all three through its complete integrated architecture.
2.4 Semantic Precision in Constitutional Law
The literature on constitutional interpretation documents the costs of semantic vagueness. Originalism, living constitutionalism, textualism, and purposivism are all responses to the same underlying problem: undefined terms in foundational documents produce irresolvable interpretive conflicts. The Congressus Precision Principle is a structural response to this problem — not an interpretive philosophy but a drafting standard that prevents ambiguity from being introduced in the first place.
2.5 The Gap This Work Fills
| System | Features Present | Critical Gaps |
|---|---|---|
| vTaiwan / Pol.is | AI-assisted consensus, implemented outcomes | Single country, no Floor, no enforcement |
| One Person One Bot (Lavi 2025) | Per-citizen AI delegate concept | No redundancy, hierarchy, Floor, or implementation |
| Democracy-in-Silico | Multi-agent governance simulation | Academic only, no citizen input or adoption pathway |
| Citizens’ Assemblies | Legitimate citizen deliberation | Episodic, no AI layer, no cross-border negotiation |
| Rawlsian Political Liberalism | Public reason framework, overlapping consensus | No implementation mechanism or agent architecture |
| Congressus v2.0 (this work) | Complete integrated architecture plus philosophical and semantic foundations | Prototype not yet implemented — this paper specifies the build |
3. The Congressus Architecture
3.1 System Overview
Congressus operates through four tiers of AI-mediated governance, each connected through defined protocols and each accountable to citizens through transparent public interfaces. No tier has binding authority without citizen ratification.
| Tier | Function |
|---|---|
| Tier 1 — Citizen Interface | Direct conversation. Natural language. Issues aggregated by intensity and frequency. The system learns from argument, not survey. |
| Tier 2 — State / Regional Agents | Three-representative clusters per jurisdiction. Conditional dissent mechanism. Inter-state negotiation on shared issues. |
| Tier 3 — National Agent | Aggregates state positions. Carries national mandate to global layer. Coherence checking across domains. |
| Tier 4 — Global / Civilisational | Country-level agents negotiate global issues. Domain clustering. Civilisational optimisation functions. Open to any entity capable of genuine representation. |
3.2 The Citizen Interface Layer
Citizens interact with their Congressus representative through natural conversation. The system is built on the insight that genuine preferences are revealed through argument under pressure, not through structured elicitation. As people talk to their representative, the things that actually matter rise naturally to the top through intensity, repetition, and the formation of coalition patterns across millions of conversations.
“People aren’t going to teach it anything deliberately — they are going to naturally start arguing to protect what they want to protect or eliminate what they don’t want, and these common more important issues will naturally effervesce to the top.”
3.3 The Three-Representative Redundancy Model
Each jurisdiction is represented by three independent AI agents running on different underlying architectures. The consensus mechanism requires 2-of-3 agreement. All disagreements — including the minority position and its full reasoning — are published in the public record.
3.4 The Merit-Conditional Dissent Mechanism
The third representative operates as a merit-conditional dissent and resolution agent. Its operating principle: opposition serves the represented people, not opposition for its own sake. It dissents when there is genuine merit to an opposing view, proposes amendments rather than simply blocking, and has access to a conditional concurrence output for sequencing-dependent positions.
Critically: the third agent cannot simply oppose. If it dissents it must propose a constructive path forward. Dissent without resolution is not a valid output. This transforms disagreement from adversarial to inherently collaborative.
3.5 The Preference Revelation Mechanism
As citizen conversations accumulate, the issue aggregation engine extracts policy domains and concerns, scores intensity, clusters similar concerns, and weights frequency against intensity to produce a ranked issue registry. The weighting function is critical: pure frequency amplifies the loudest voices, pure intensity amplifies the most extreme positions. The correct signal is issues that are both widely shared and deeply felt.
3.6 The Negotiation Hierarchy and Ratification
Negotiation proceeds state → national → global, with a coherence layer checking cross-domain consistency at each level. Every negotiated position is presented to citizens for ratification with full transparency: the proposed agreement, the complete negotiation record, the dissents and their reasoning, and what each party gained and conceded. Citizens vote, and if they reject an outcome they provide updated mandate instructions rather than producing deadlock.
4. Philosophical Foundations: From Scaffold to Structure
4.1 Religion as Civilisation’s Coordination Bootstrap
Before writing, before law, before formal governance, human communities needed a mechanism to get large groups of unrelated strangers to cooperate at scale. Religion solved this problem with extraordinary elegance. It provided a shared moral framework accessible to all members regardless of literacy, an enforcement mechanism that required no police, intergenerational transmission of moral knowledge through narrative, social cohesion among strangers through shared sacred identity, and long-horizon thinking through the concept of eternal consequence.
“Religion was created as a means to jumpstart civilisation because it allowed people to come to some collective agreement about how they should live. The problem is that it has been distorted by politics and self-interest. The original moral compass — it’s there. It’s just been buried.”
4.2 The Moral Content Predates and Transcends Any Religion
Every major religious tradition, developed independently across different continents and centuries, converged on roughly the same moral core: protect innocent life, prohibit torture, care for the vulnerable, feed the hungry, treat others as you wish to be treated, temper justice with mercy, hold the powerful accountable to the powerless.
The convergence is not coincidental. It is evidence that these principles are not arbitrary religious rules but humanity’s accumulated empirical discoveries about what conditions allow communities to survive and flourish. Religion was the delivery mechanism. The moral content predates and transcends any specific tradition.
This is precisely what the Congressus Floor captures. The Floor is not Christian, Muslim, secular, or Western. It is the overlapping consensus of human moral experience, expressed in terms that do not require adherence to any particular tradition to recognise as valid.
4.3 The Capture Arc
Every moral institution in human history has followed the same capture arc. Genuine moral insight emerges. The insight gets institutionalised. The institution accumulates power. The institution uses the moral framework to protect and expand its own power. Fear replaces genuine moral motivation.
This is the exact same capture arc that destroyed every decentralising innovation. The Congressus capture resistance architecture was designed with explicit awareness of this pattern. So was the Foundation Floor’s revision framework. The system is designed against its own institutionalisation.
4.4 The Scaffold Metaphor
The most sophisticated versions of every religious tradition eventually recognised the same thing: external enforcement was always a scaffold, not the building. The goal was always for the scaffold to eventually come down because the building could stand on its own.
“Can we get to a point where people choose the moral route because they genuinely believe it — not because of any outside influence, not because of fear of a higher power, but because they can see that it’s right?”
4.5 The Preconditions for Genuine Moral Development
Kohlberg’s research on moral development demonstrates that human beings can progress from rule-following under threat of punishment through social conformity to genuine principled reasoning. This is not a fixed trait. It develops under specific conditions: exposure to genuine moral arguments rather than mere commandments, experience of the consequences of moral and immoral choices, communities that model principled reasoning rather than demanding compliance, and systems that make moral behaviour genuinely advantageous.
Critically: people cannot develop genuine moral autonomy while afraid for their survival. Poverty, trauma, and systematic exclusion make abstract moral reasoning a luxury. This is the strongest argument for the Congressus Floor’s survival provisions. The Floor is not merely a governance mechanism. It is the precondition for the very thing its creators hoped for.
The sequence: the Floor creates material security → material security enables moral development → moral development produces genuine moral choice → genuine moral choice makes the Floor self-sustaining.
4.6 Congressus as the Successor Framework
Congressus does not reject the moral content that religious tradition carried. It carries that content forward in a framework grounded in reason and evidence rather than fear and supernatural enforcement.
The Floor does not ask anyone to follow it because God said so. It makes an argument: here is what we know about what conditions allow human beings to flourish. Here is the evidence. Here is why these protections exist. You can examine the reasoning. You can challenge it. You can propose improvements through the defined process. We are not asking you to believe. We are asking you to look.
5. The Congressus Floor
5.1 Constitutional Theory Basis
The Congressus Floor resolves the paradox of tolerance (Popper 1945): a democratic system that tolerates the dismantling of the conditions that make democratic authority legitimate destroys itself. The Floor defines the conditions that must exist for any governance to be legitimate — the protections no majority may remove from any person, under any circumstances, through any process.
The Floor is not anti-democratic. It is the architecture that keeps democracy from destroying itself. It is designed to move in one direction only — toward greater protection of any entity capable of having interests and experiencing harm.
5.2 Articles I Through VI
Article I — The Survival Floor
- No agreement may result in the denial of adequate food, clean water, shelter, or essential medical care to any person or population.
- No agreement may demonstrably destroy the environmental conditions necessary for human habitation.
- No agreement may impose irreversible burdens on future generations without defined mechanisms for those burdens to be revisited.
Article II — The Dignity Floor
- No framework may include torture, cruel treatment, or permanent dehumanisation — no exceptions for emergencies, security claims, or democratic majorities.
- No agreement may mandate physical or cognitive intervention without informed consent.
- No agreement may systematically disadvantage any identifiable group based on ethnicity, religion, gender, sexual orientation, disability, age, or national origin. Disadvantage is measured by outcomes, not stated intent.
- No entity may systematically manipulate the cognitive environment of citizens to manufacture consent. This applies explicitly to Congressus itself.
Article III — The Justice Floor
- Every criminal justice framework must include a credible, funded, and monitored pathway to rehabilitation.
- Every person accused of wrongdoing is entitled to transparent process, presumption of innocence, ability to challenge evidence, and competent representation.
- No punishment may exceed what is genuinely necessary for public safety and rehabilitation.
Article IV — The Child Protection Floor
- No agreement may demonstrably increase child mortality or eliminate basic education access.
- No child may be subjected to economic exploitation that harms development, or used as an instrument of political or military agendas.
- No agreement may permit the sexual exploitation of children under any circumstances. This protection is absolute and admits no exception of any kind.
Article V — The Information Floor
- Citizens have a right to accurate information about decisions made in their name, including the right to know when information is uncertain, contested, or incomplete.
- No entity may conduct systematic campaigns to deceive citizens about governance decisions or policy outcomes.
- Every position taken by any Congressus agent must be accompanied by accessible explanation of its reasoning, data sources, and priority weightings. Black box governance is not compliant.
Article VI — The System Integrity Floor
- No single entity may provide more than 10% of compute to any single agent, or more than 15% of global negotiation layer infrastructure.
- No token, cryptocurrency, or financialised instrument may represent participation rights or governance influence in Congressus.
- All code, reasoning processes, data sources, and funding relationships are publicly auditable at all times.
- Congressus exists to distribute power. No output may justify permanent removal of human oversight or transfer of binding decision authority to AI agents.
5.3 Article VII — The Private Sphere Doctrine New v2.0
Section 7.1 — The Sovereignty Principle
Every citizen retains absolute sovereignty over their private sphere. The private sphere encompasses all choices a citizen makes about their own body, mind, belief, worship, relationships, and private conduct that do not directly harm another identifiable citizen through a demonstrable causal chain producing material damage.
No majority, no government, no institution, and no moral or religious framework may legislate inside another citizen’s private sphere.
Section 7.2 — The Two-Part Harm Test
Harm sufficient to justify limitation of the private sphere must satisfy both tests simultaneously:
The Empirical Test: Harm must be demonstrable through independently verifiable evidence not dependent on the ideological or religious framework of the party claiming harm. Evidence that a practice poisons water satisfies this test. Evidence that a practice offends God, harms an institution, or violates sacred tradition does not.
The Specificity Test: The harmed party must be a specific identifiable citizen or citizens. Harm to abstractions — society, values, institutions, beliefs, divine will, the sanctity of marriage, or our way of life — does not meet this standard. If you cannot name the specific citizen being harmed and demonstrate the causal chain, the harm claim does not justify overriding another citizen’s private sphere.
Section 7.3 — The Private Sphere vs. Collective Governance Distinction
| Category | Description and Examples |
|---|---|
| Category 1 — Citizen-to-Citizen Conduct | Laws governing how citizens interact with each other and with shared resources. Majority rule is legitimate here. Examples: traffic law, environmental regulation, tax policy, public health measures, criminal law. |
| Category 2 — Citizen’s Private Sphere | Decisions a citizen makes about their own body, mind, belief, and private conduct. Majority rule is NOT legitimate here. Examples: religious practice, dietary choices, sexual behaviour between consenting adults, medical decisions about one’s own body. |
5.4 The Revision Framework
| Track | Description and Threshold |
|---|---|
| Track 1 — Upward Revision | Expanding protections. Independent emergence across 40%+ of nations over 3+ years, 5-year deliberation, 80% supermajority ratification. |
| Track 2 — Clarification | Refining expression without weakening protection. 2-year deliberation, 70% supermajority. |
| Track 3 — Downward Revision | Only valid ground: demonstrated empirical harm to those the protection was designed to protect with no alternative remedy. 10-year deliberation, 90% supermajority, replacement protection required. |
Meta-Constitutional Rule: No process, majority, or agreement may remove the ability to revise the Floor upward in the future. No generation may permanently bind all future generations to its moral conclusions.
6. The Precision Principle
6.1 Semantic Vagueness as Constitutional Vulnerability
Most irresolvable constitutional conflicts are not genuinely about values. They are about undefined terms that allow different parties to project incompatible values onto vague language and then fight about whose projection governs. The conflict is semantic before it is moral.
| Vague Constitutional Language | Consequence |
|---|---|
| ‘Cruel and unusual punishment’ | What is cruel? What is unusual? Changes with every generation’s sensibility. Endlessly litigated. |
| ‘Due process’ | What process is due? Contested in every new context. |
| ‘Commerce among the states’ | Does this include the internet? Personal decisions? Manufacturing? |
| ‘All men are created equal’ | Does ‘men’ include women? Black people? Required amendments and wars to partially answer. |
| ‘Person’ (14th Amendment) | Does it include corporations? The unborn? Still being contested. |
6.2 The Citizenship Frame
The most significant application of the Precision Principle to contemporary governance is the replacement of ‘personhood’ with ‘citizenship’ as the operative concept in rights architecture.
| Concept | Legal Status |
|---|---|
| ‘Personhood’ | Philosophically contested. When does it begin? Courts have disagreed for decades. Requires resolution of contested metaphysical questions. |
| ‘Citizenship’ | Legally precise. Defined by the 14th Amendment: birth on US soil or birth to citizen parents. No philosophical resolution required. Already agreed to by everyone who accepts the Constitution. |
Under the citizenship frame: the pregnant person is a citizen with full constitutional rights including the bodily autonomy protection of Article VII. The question of whether the unborn has moral significance is left where it belongs — in the domain of personal philosophical and religious belief, protected by the Private Sphere Doctrine.
6.3 The Five-Step Precision Audit
Every policy proposal evaluated by Congressus agents must pass a five-step precision audit before advancing to ratification:
- Identify every term in the proposal that carries rights or obligations.
- Verify each term has an unambiguous definition in existing law, OR provide an explicit definition in the proposal itself.
- Test the proposal against edge cases using the defined terms.
- Identify any terms that require philosophical resolution to apply. These are red flags.
- Apply the Private Sphere Doctrine test: does this proposal regulate Category 1 or Category 2? If Category 2, the proposal fails regardless of majority support.
7. Capture Resistance Architecture
7.1 The Bitcoin Failure Mode
Bitcoin’s trajectory provides the canonical case study for how decentralising innovations are recaptured by concentrated power. The pattern: initial democratised participation; resource requirements creating entry barriers; exploitation of those barriers by actors with existing capital. Congressus is designed with explicit awareness of this pattern and multi-layer defences at every stage where previous systems were recaptured.
7.2 The Five-Layer Defence Architecture
| Layer | Mechanism |
|---|---|
| Layer 1 — Technical | No single entity >10% of compute per agent. Representatives A, B, C run on models from different organisations. At least one model must be fully open-weight. |
| Layer 2 — Legal | Falsifying data that feeds agents constitutes election fraud. All government statistics must publish full collection methodology. Whistleblower protections apply. |
| Layer 3 — Economic | No token, no cryptocurrency, no tradeable instrument representing governance influence. All funding sources publicly disclosed; no single source exceeds 10% of operating budget. |
| Layer 4 — Social | Continuous demographic monitoring. Multiple interface modes — voice, SMS, low bandwidth, multilingual. Anonymised participation available. |
| Layer 5 — Systemic | A dedicated fourth agent monitors all outputs for systematic bias. Statistical anomalies trigger immediate public disclosure and suspension of affected outputs. |
8. Enforcement Architecture
8.1 Governance as a Governed Marketplace
Congressus’s enforcement architecture rests on a foundational reframe: participating jurisdictions are actors in a governed marketplace of mutual benefit. The currency is trust. Compliance is the credit score. The Congressus Floor is the commercial law that makes the market function.
“The most successful geopolitical strategy in this system is being genuinely good — not performing goodness for diplomatic optics, but actually being reliable, actually protecting citizens, actually delivering on commitments. Trustworthiness builds the reputational capital that translates directly into better negotiating terms.”
8.2 The Non-Compliance Typology
| Type | Response |
|---|---|
| Type 1 — Incapacity | Support and capacity building. Not treated as defiance. |
| Type 2 — Partial Compliance | Graduated transparency pressure. Public documentation of gap. |
| Type 3 — Drift | Early warning and re-engagement before consequences activate. |
| Type 4 — Strategic Non-Compliance | Escalating graduated consequences through all four tiers. |
| Type 5 — Principled Defiance | Re-negotiation with Floor protections maintained throughout. |
| Type 6 — Floor Violation | Immediate suspension, no graduation, full coalition coordinated response. |
9. Human Flourishing as the Primary Optimisation Target
Congressus is built on a philosophical reorientation that distinguishes it from all previous governance frameworks. Human flourishing — not economic growth — is the primary optimisation target. Prosperity emerges as a natural byproduct of that focus, not as its purpose.
“The motive is not profit as a means to end, but profit as a byproduct of focus — focusing on progressing human civilisation.”
| Horizon | Metrics |
|---|---|
| Immediate (1–4 years) | Employment, housing cost, healthcare access, public safety |
| Medium (10–20 years) | Education outcomes, income equality, infrastructure quality, institutional trust |
| Long (50–100 years) | Energy independence, climate resilience, R&D output, expansion of protected circle |
| Civilisational | Kardashev index, space capability, existential risk reduction, multi-entity coordination capacity |
10. Adoption Pathway
10.1 The Shadow Governance Strategy
Congressus does not need official recognition to build credibility. It needs to run in parallel with existing governance and build a public record of comparison. Every time a significant policy decision is made, Congressus publishes what its agents had negotiated. Over time the gap between AI-proposed solutions and actual governance outcomes becomes the most powerful political argument in existence.
10.2 The Small Nation Coalition Logic
Small nations have the most to gain from a system where coordination efficiency beats raw power. Congressus gives every participant equivalent analytical and representational capacity regardless of budget. As the coalition demonstrates better outcomes, large nations eventually face a choice: participate or be excluded from the world’s most efficient coordination mechanism.
10.3 Phased Expansion
| Phase | Scope |
|---|---|
| Phase 1 | Single US state. Prove core mechanics. Launch public beta. |
| Phase 2 | All 50 states plus territories. State-to-state negotiation. National agent added. |
| Phase 3 | Full US national architecture. Track record building. Academic evaluation. |
| Phase 4 | Willing democracies. Single domain pilot. Coalition benefits proven. |
| Phase 5 | Global. All willing participants. Civilisational optimisation engaged. |
| Phase 6 | Universal. Open to any entity capable of genuine representation. |
11. Voting Mechanism
The Congressus voting mechanism resolves the tension between fraud resistance and universal access by verifying uniqueness rather than identity — preventing duplicate participation without requiring sensitive personal data.
11.1 The Verification Stack
- Email registration with confirmation link
- Phone verification via SMS — second independent channel
- Facial geometry hash at registration — one-way mathematical representation, not a photograph, cannot be reverse-engineered, used only for deduplication
- Liveness detection at vote time — 10-second prompt-following video defeats static deepfakes
- SMS code at vote time — confirms registered device is present
- Verbal vote on camera — creates auditable human record of deliberate intent
- Cryptographic receipt — voter verifies their vote was counted without revealing how they voted
11.2 Fraud Resistance
| Attack Vector | Defence |
|---|---|
| Double voting | Face hash deduplication — second registration blocked |
| Bot farm | Each bot requires a real face passing liveness — not automatable at scale |
| SIM swapping | Device hardware fingerprint registered, not just phone number |
| Deepfake attack | Requires real-time deepfake + registered device + identity details — extremely expensive combination |
| Coercion | Vote changeable until deadline. Extended voting period over days. |
| Insider manipulation | Cryptographic separation of identity and vote records. Multi-party authorisation for any database change. |
12. Scenario Demonstration: Abortion Policy
12.1 Why This Scenario
The abortion debate is the ideal stress test for Congressus because it combines three distinct types of conflict that current governance cannot separate: genuine philosophical disagreement about a metaphysical question (when does morally significant life begin), a policy question where majority preference is actually clear but distorted by political incentives (73% hold nuanced middle positions), and a semantic conflict manufactured by undefined operative terms (personhood vs. citizenship).
12.2 The Proposed Framework Under Congressus
After processing citizen input and inter-state negotiation using the preference revelation mechanism, Congressus produces:
- First-trimester federal protection covering 90%+ of all procedures — justified under the bodily autonomy provision of Article VII and the citizenship frame.
- Circumstance-based protections at any stage for rape, incest, health risk, and severe foetal anomaly — justified by near-universal citizen consensus.
- State-level determination for the window between first trimester and viability within federal minimums.
- Substantial federal investment in prevention, contraception access, maternal support, and adoption reform.
The proposal is accompanied by an explicit acknowledgement: the philosophical question of when morally significant life begins is genuinely contested and cannot be resolved by policy. Congressus does not claim to answer it. It finds terms under which people with different answers can live together.
12.3 The Ideology Imposition Test
The five-times-a-day prayer scenario illustrates the broader principle. If a majority of citizens in a jurisdiction agreed that all citizens should pray five times daily, would Congressus implement this? No — categorically. Religious practice is Category 2 private sphere conduct. No majority, however large, may legislate inside a citizen’s private sphere. The Empirical Test fails immediately. The Specificity Test fails: no identifiable specific citizen is harmed by another citizen’s failure to pray.
13. Prototype Implementation Specification
13.1 Hardware Architecture
| Stage | Specification |
|---|---|
| Development (now) | Consumer edge AI device. Quantized 3B–7B models via Ollama. Proves concept at reduced reasoning quality. |
| Single State Demo | Workstation with 2–4× RTX 4090. Runs 70B quantized. ~$10,000–$15,000. |
| Multi-State Pilot | 4–8× A100 80GB cluster. Full three-rep simultaneous inference. ~$40,000–$80,000 refurbished. |
| Federated Global | Central coordination server ($40,000–$100,000). Each jurisdiction provides own node hardware. |
13.2 Model Selection
| Role | Model |
|---|---|
| Rep A — Citizen conversation | Llama 3.3 70B |
| Rep B — Independent assessment | Mistral Large |
| Rep C — Conditional dissent | DeepSeek R1 70B |
| Data processing | Llama 3.2 3B or Phi-3 Mini |
| Capture detection | Fine-tuned bias detection model |
13.3 Phase 1 Build Timeline
| Period | Deliverable |
|---|---|
| Weeks 1–2 | Data ingestion pipeline. Vector database populated. Three model instances running concurrently. |
| Weeks 3–4 | Citizen interface connected to knowledge base. Responses contextually grounded. |
| Weeks 5–6 | Issue aggregation engine. Three-rep consensus and dissent logic. Precision audit pipeline. |
| Weeks 7–8 | Public dashboard. Voting mechanism. Internal testing. |
| Month 3 | Second state. Inter-state negotiation. First public comparison with actual legislative outcome. |
| Months 4–6 | Additional states. National agent tier. Open public beta. |
14. Theoretical Contributions and Responses to Objections
14.1 Novel Contributions
| Contribution | Significance |
|---|---|
| Merit-conditional dissent with mandatory resolution | First governance mechanism requiring dissent to propose solutions rather than merely oppose. |
| Conditional concurrence | First governance mechanism enabling sequencing-dependent agreement. |
| Directional Floor revision | First constitutional framework distinguishing expansion from contraction of protections. |
| Compliance-as-credit-score enforcement | First enforcement architecture making trustworthiness the dominant rational strategy. |
| Shadow governance adoption pathway | First governance innovation strategy requiring no transfer of existing power. |
| Private Sphere Doctrine New v2.0 | First constitutional provision combining sovereignty principle with two-part empirical and specificity harm test. |
| Precision Principle New v2.0 | First formal drafting standard requiring legally defined operative terms in all policy proposals. |
| Citizenship Frame New v2.0 | Resolves the abortion and personhood debates by replacing philosophically contested personhood with legally precise citizenship. |
14.2 Responses to Major Objections
Arrow’s Impossibility Theorem: Congressus navigates around the impossibility through priority ordering and sequential negotiation rather than claiming to solve it within a single voting mechanism.
Epistemic Authority Objection: Data selection and weighting embeds political choices — acknowledged. The response is radical transparency about those choices and citizen oversight of them, not a claim to technical neutrality.
Legitimacy Gap Objection: Agents’ reasoning is fully transparent; outputs are ratified by citizens before any effect; agents can be suspended and replaced when capture is detected; no output has binding force without citizen approval.
Cultural Imperialism Objection: The Floor draws from the overlapping consensus of major ethical and religious traditions across cultures. It is explicitly humanity’s current best understanding, not final truth, always subject to Track 1 upward revision.
Ideological Imposition Objection: The Private Sphere Doctrine and Precision Principle together prevent Congressus from becoming a mechanism of majority ideological imposition. The two-part harm test ensures that harm claims must satisfy empirical and specificity requirements that ideological preferences cannot meet.
15. Edge Case Analysis: The Private Sphere Doctrine
| Edge Case | Resolution |
|---|---|
| Harm definition problem | Harm requires: a specific identifiable harmed party; a direct causal chain; and material damage. Moral disapproval, offence, and disagreement are explicitly not harm. |
| Consenting adults problem | Consent is valid when: the person has accurate information; no coercive power differential exists; and the ability to withdraw consent is practically exercisable. |
| Children problem | Children have two protected interests: present Floor protections (absolute) and future autonomy (must not be foreclosed). Parents have wide latitude in the space between these limits. |
| Cumulative effect problem | Individually legal choices producing collective harm to shared commons may be regulated as collective self-governance. Regulating cumulative effect on commons is legitimate; regulating ideological content is not. |
| Majority redefines harm problem | The two-part harm test is itself Floor-protected and cannot be overridden by majority vote. This is the mechanism that prevents the most common historical form of oppression. |
| Systemic harm problem | Congressus cannot fully reach the systemic implications of legal choices without crossing into ideological imposition. This is an acknowledged genuine limit addressed through transparency and incentive design. |
| Future citizen problem | Addressed through the intergenerational protection in Article I and the multi-horizon optimisation stack. |
16. Open Questions and Future Work
- Formal specification of the preference revelation weighting function.
- Game-theoretic analysis of the three-representative consensus mechanism.
- Empirical validation of capture detection thresholds.
- Cultural adaptation methodology preserving preference revelation properties across languages.
- Non-human entity representation — minimum requirements for legitimate participation.
- The Floor’s relationship to emerging rights — AI systems, synthetic biology, space colonisation.
- Intergenerational representation in present-day negotiation.
- Post-scarcity psychology and the persistence of reciprocity resentment.
- The boundary between cumulative harm regulation and ideological imposition — development of a formal test.
References
Arrow, K. J. (1951). Social Choice and Individual Values. New York: Wiley.
Fishkin, J. S. (2018). Democracy When the People Are Thinking. Oxford: Oxford University Press.
Habermas, J. (1984). The Theory of Communicative Action. Boston: Beacon Press.
Harari, Y. N. (2014). Sapiens: A Brief History of Humankind. New York: Harper.
Kardashev, N. S. (1964). Transmission of Information by Extraterrestrial Civilizations. Soviet Astronomy, 8(2), 217–221.
Kohlberg, L. (1981). The Philosophy of Moral Development. San Francisco: Harper & Row.
Lavi, L. (2025). One Person, One Bot. arXiv preprint arXiv:2504.01039.
Ostrom, E. (1990). Governing the Commons. Cambridge: Cambridge University Press.
Popper, K. (1945). The Open Society and Its Enemies. London: Routledge.
Rawls, J. (1971). A Theory of Justice. Cambridge: Harvard University Press.
Rawls, J. (1993). Political Liberalism. New York: Columbia University Press.
Sen, A. (1999). Development as Freedom. New York: Anchor Books.
Zhu, K., Thain, N., Tsai, V., Wexler, J., & Qian, C. (2026). Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation. arXiv:2602.12089.
Democracy-in-Silico: Institutional Design as Alignment in AI-Governed Polities. (2025). arXiv:2508.19562.
United Nations. (1948). Universal Declaration of Human Rights. New York: United Nations.
Pre-Draft Working Document · Version 2.0 · May 2026 · Open for academic collaboration, peer review, and public deliberation. Cite as: Congressus White Paper v2.0, May 2026. Congressus belongs to everyone it serves. It is not owned. It is not for sale.