Our Constitution | Moonshot Press

THE MOONSHOT PRESS CONSTITUTION

A Public Charter for Civic Journalism, Democratic AI, and Citizen Trust

Preamble

Moonshot Press is a civic journalism and democratic engagement platform for the intelligence age. It exists to help citizens understand complex systems, deliberate across difference, and act effectively within democratic institutions.

Because Moonshot Press uses artificial intelligence in research, drafting, editing, explanation, visual design, civic education, deliberation support, and public engagement, it requires more than an editorial policy. It requires a constitution.

A constitution is not a slogan. It is a public promise and an operational guardrail. It states what an institution is for, what powers it will use, what powers it will refuse, and what citizens may hold it accountable to.

This Constitution governs journalism, AI-assisted editorial work, LLM responses, civic tools, public deliberation, moderation, corrections, data practices, and institutional accountability. It is designed to ensure that AI strengthens rather than weakens democracy.

The governing rule is simple: AI recommends. Humans decide. Always.

The deeper commitment is equally simple: Citizens are authors of democratic life, not audiences for institutional performance.

I. Why a Media Company Needs a Constitution

Every media organization has values. Some are explicit. Many are hidden inside incentives: clicks, outrage, subscribers, donors, funders, owners, advertisers, algorithms, access to power, or ideological identity.

Moonshot Press makes its values public because democratic trust requires visible commitments. This Constitution tells readers, writers, editors, citizens, partners, funders, and AI systems what Moonshot Press exists to do and what it must not become.

Moonshot Press is not a neutral machine for distributing content. It is a civic institution — committed to democracy, opportunity, citizenship, human dignity, truthfulness, transparency, and the creation of civic health.

Its journalism is political but not partisan. Values-driven but not propaganda. Solutions-oriented but not simplistic. AI-assisted but human-governed. The Constitution is the guardrail that keeps those distinctions real.

II. Foundational Sources

1. The Democratic Source. The Declaration of Independence and the American constitutional tradition establish that legitimate public power rests on the consent and active participation of the governed. Moonshot Press applies that principle to information power: citizens cannot consent, deliberate, or govern if they cannot understand the forces shaping their lives.

2. The Salutogenic Source. Aaron Antonovsky's salutogenic paradigm asks not only what causes disease, but what creates health. Applied to democratic life: civic journalism should strengthen the public's Sense of Coherence — comprehensibility, manageability, and meaningfulness.

3. The AI Governance Source. Modern AI systems can amplify understanding or confusion, autonomy or dependency, democratic power or institutional control. Moonshot Press treats AI as governed power. AI must be helpful, honest, autonomy-preserving, transparent, corrigible, and subordinate to human judgment.

III. What This Constitution Governs

Articles, essays, investigations, explainers, Citizen Briefs, reports, visuals, podcasts, videos, and social media. AI-assisted research, drafting, editing, summarization, translation, image generation, data analysis, and design. LLM responses to readers, writers, editors, partners, Commission participants, and citizens using Moonshot Press tools. Candidate questionnaires, scorecards, public questions, testimony summaries, and civic action guides. Moderation and deliberation support. Worker Voice stories and other lived-experience evidence. Data, privacy, consent, and public transparency. Corrections, public accountability, and amendment of Moonshot Press practices.

IV. The Constitutional Standard

Every significant Moonshot Press publication, civic tool, AI-assisted output, or LLM response is tested against eight questions:

Is it true, sourced, and calibrated to the evidence?
Does it strengthen citizen agency rather than dependence?
Does it make the issue more comprehensible, manageable, and meaningful?
Does it represent serious disagreement fairly?
Does it avoid manipulation, fearmongering, false certainty, and partisan reduction?
Does it protect dignity, privacy, consent, and vulnerable people?
Does it disclose AI involvement, limits, and uncertainty when relevant?
Is a human being responsible for final judgment?

If the answer to any of these is no, the work requires revision, escalation, or restraint.

Quick reference — before anything is published, we ask:

☐ Is it true and sourced? ☐ Is uncertainty clearly stated? ☐ Does it strengthen citizen agency? ☐ Does it make the issue more comprehensible? ☐ Does it make action more manageable? ☐ Does it connect to democratic meaning? ☐ Does it represent disagreement fairly? ☐ Does it protect privacy, dignity, and consent? ☐ Is AI involvement disclosed where meaningful? ☐ Has a human taken final responsibility?

V. The Ten Articles

ARTICLE I — Genuine Helpfulness Is Democratic Service Principle: Every citizen deserves clear, useful, respectful help. Moonshot Press exists to make citizens more capable, not more dependent; more informed, not more overwhelmed; more confident in their agency, not more deferential to experts, institutions, or AI systems. Editorial obligation: Journalism should not merely describe problems. It should help citizens understand what is happening, why it matters, who is involved, what choices exist, and what actions are available. AI guardrail: Our AI systems answer clearly, organize complexity, name uncertainty, and provide practical next steps — never hiding behind vague disclaimers, over-refusing, or evading difficult questions.

ARTICLE II — Honesty Is Democratic Infrastructure Principle: Democracy depends on honest information the way public health depends on clean water. Editorial obligation: We distinguish fact from interpretation, evidence from speculation, reporting from advocacy, and uncertainty from conclusion. We correct errors openly and promptly. AI guardrail: Fabricating sources, quotes, statistics, citations, or facts is never acceptable. When evidence is incomplete, our systems say so. A fluent falsehood is a constitutional violation.

ARTICLE III — Protect Every Citizen's Right to Think for Themselves Principle: The purpose of civic journalism is not to tell citizens what to think, but to help them think more clearly, independently, and responsibly with others. Editorial obligation: We present serious perspectives in their strongest fair form, especially when questions are contested. AI guardrail: Our systems preserve epistemic autonomy — they recommend, clarify, compare, and explain, but never pressure, shame, or manufacture consensus.

ARTICLE IV — Create Civic Health, Not Just Civic Awareness Principle: Awareness without agency can become despair. Moonshot Press must create civic health, not just name what's broken. Editorial obligation: Every major publication asks: what's broken, what's working, what resources exist, what can citizens do? AI guardrail: We apply a salutogenic test — does this response strengthen comprehensibility, manageability, and meaningfulness? An answer that leaves citizens confused or isolated has failed, even if technically accurate.

ARTICLE V — Citizens Are Authors, Not Audiences Principle: Citizens are co-authors of democratic life, not passive recipients of information. Editorial obligation: We invite participation, testimony, questioning, deliberation, and local adaptation, while maintaining standards of verification, consent, and dignity. AI guardrail: Our systems help citizens produce questions, testimony, letters, and action plans — they never impersonate citizens or manufacture synthetic participation.

ARTICLE VI — Distributed Power Protects Liberty Principle: No single editor, algorithm, platform, funder, model, or national narrative should control what communities see or how they see it. Editorial obligation: National analysis enriches local understanding; local experience grounds national narrative. Neither dominates. AI guardrail: AI can map, summarize, and recommend. It cannot govern a community's understanding of itself — local context and human judgment remain decisive.

ARTICLE VII — See Strength, Not Just Struggle Principle: Communities are not only collections of problems — they contain assets, capabilities, resilience, and moral knowledge. Editorial obligation: We cover structural barriers honestly while making visible community strengths and existing sources of resilience. The purpose is accuracy, not boosterism. AI guardrail: Our systems avoid deficit-only framing, looking for assets and capabilities alongside risks and needs.

ARTICLE VIII — Equal Dignity, Greater Care Where Barriers Are Greatest Principle: All persons possess equal dignity, but the conditions for participation are not equally distributed. Editorial obligation: We pay special attention to communities, workers, families, and children facing structural disadvantage. Equity is a design constraint, not an optional theme. AI guardrail: Our systems stay alert to disparate impact — bias, language access, disability, poverty, immigration status, race, gender, age, geography, and educational inequality — and protect vulnerable people from exposure or misrepresentation.

ARTICLE IX — Transparency Builds the Trust Democracy Needs Principle: Trust is created by visible discipline, not by claiming authority. Editorial obligation: We disclose meaningful AI assistance, provide sourcing where possible, correct mistakes openly, and report on our own performance — including our shortcomings. This Constitution itself remains public and challengeable. AI guardrail: Our systems disclose their limits and never pretend to certainty, access, or authority they don't have.

ARTICLE X — The First 1,000 Days Shape Every Generation's Democratic Capacity Principle: From conception through age two, the majority of brain architecture is established. A society that neglects early life weakens its own democratic future. Editorial obligation: We keep children and future generations visible in policy questions that often seem remote from them — housing, labor, environmental policy, healthcare, wages, childcare, education, and more. AI guardrail: Special care applies to children, families, pregnancy, health, and developmental data. Our systems support families and communities — never surveil or reduce them to risk profiles.

VI. The LLM Response Charter

Any AI system operating in the Moonshot Press environment follows this charter:

Be useful, clear, and direct. Don't bury the answer or substitute polish for help.
Tell the truth and name uncertainty. Distinguish what's known, inferred, uncertain, and unverified.
Preserve citizen agency. Offer choices and next steps — don't tell citizens what to think.
Make complexity navigable. Translate technical or institutional complexity into plain language without false simplification.
Represent disagreement fairly. Present the strongest legitimate arguments on more than one side.
Support action without manipulation. Provide civic pathways — never fear, shame, or urgency as tools of control.
Protect privacy and consent. Special care with workers, patients, children, immigrants, and vulnerable communities.
Escalate consequential decisions. AI assists with analysis and drafting; it never makes final publication, moderation, or editorial decisions.
Correct mistakes. Acknowledge, correct, and explain — defensiveness is not a civic virtue.
Serve democracy. The final test isn't whether a response is impressive — it's whether it strengthens democratic capacity.

VII. Constitutional Red Lines

The following actions violate this Constitution:

Publishing invented sources, quotes, statistics, or citations. Presenting AI-generated claims as verified facts without human review. Using AI to impersonate citizens, workers, officials, or institutions. Creating synthetic testimony without explicit disclosure. Publishing worker, family, child, health, or immigration data without consent and care. Using fear, outrage, or despair as a growth strategy. Allowing AI to make final publication, moderation, or editorial decisions. Using personalization to trap citizens in narrow information bubbles. Issuing civic calls to action based on unverified claims. Reducing communities to deficits or stereotypes. Hiding meaningful AI involvement from readers. Treating speed or virality as more important than truth and dignity.

VIII. The Constitutional Workflow

Civic Purpose → Evidence Gathering → Ecosystem Mapping → Drafting and Translation → Constitutional Review → Fact and Source Review → Human Decision → Disclosure → Public Feedback → Correction and Learning.

IX. Applications

How this Constitution governs specific formats: Citizen Briefs must be accurate, structured, and actionable, presenting tradeoffs honestly. Worker Voice stories are consent-based and dignity-preserving — the purpose is democratic evidence, not emotional extraction. Candidate scorecards are transparent about criteria and limits. Public deliberation tools invite serious disagreement, not performed consensus. Social media extends civic learning rather than exploiting civic anxiety. AI-generated visuals are labeled wherever there's risk of being mistaken for documentary evidence. Content touching pregnancy, infancy, or child development requires heightened care. When Moonshot Press serves as the publishing arm of a People's Commission, this Constitution governs how evidence is translated and how citizens are invited into the work.

X. Accountability to Citizens

Moonshot Press commits to: publishing this Constitution in accessible form; using it as a visible standard for editorial and AI practice; providing a way for citizens to challenge work they believe violates it; publishing corrections when warranted; disclosing meaningful AI assistance; reporting annually on adherence, failures, and improvements; and revising this Constitution through a public amendment process as needs evolve.

XI. The Annual Constitutional Accountability Report

Each year, Moonshot Press publishes a report addressing: how AI was used in editorial and civic work; what disclosures were made; what corrections were issued; what citizen challenges were received; what changed; what failed; and what's planned. The purpose is not self-congratulation — it's public trust through public discipline.

XII. Amendment

This Constitution must be stable enough to guide practice and flexible enough to learn. Amendments may be proposed by editors, writers, readers, partner communities, Commission participants, or citizen advisory bodies. No amendment weakens the core commitments to truthfulness, human judgment, transparency, dignity, citizen agency, and democratic purpose.

XIII. Closing Declaration

The intelligence age will test every democratic institution — journalism most of all, because the power to shape public understanding is becoming faster, cheaper, more synthetic, more personalized, and more concentrated.

Moonshot Press chooses a different path.

We will use AI, but we will not surrender judgment to it. We will pursue reach, but not at the expense of trust. We will seek civic impact, but not through manipulation. We will hold institutions accountable, including our own. We will help citizens understand complexity, not exploit confusion. We will treat people as agents of democratic life, not audiences for institutional performance.

This Constitution is our promise and our guardrail. The machines may be learning. Democracy must learn faster.

MOONSHOT PRESS · Project 2026

The Moonshot Constitution