Governance
Governance
1. Overview
Governance defines oversight, neutrality, ethical principles, and organisational integrity.
It ensures that FactHarbor operates transparently, without bias, and in line with its mission.
FactHarbor is intended as a long-term public knowledge infrastructure.
Its governance model must ensure:
- integrity against manipulation
- transparency of reasoning and code
- longevity beyond any single founder
- sustainability via realistic financing
- openness without compromising trust
- legal compliance under relevant frameworks (e.g. Swiss, EU, and US law, as applicable)
Governance is responsible for:
- mission alignment and ethical oversight
- safeguarding neutrality and independence
- defining and supervising decision processes
- ensuring that organisational structures remain fit for purpose as the project evolves
2. Charter
*Mission (governance focus):*
Protect neutrality, transparency, nonprofit accountability, and ethical integrity across all FactHarbor activities.
*Legal foundation (high-level):*
FactHarbor’s governance is designed to operate within the constraints of:
- nonprofit law and applicable registration requirements
- licensing obligations (open-source and open-data)
- financial reporting and transparency duties
- data protection and information-security regulations
Details of the concrete legal form (e.g. association, foundation, other nonprofit structures) may evolve over time.
The core commitment is that governance must always protect the mission and public-interest character of FactHarbor.
3. Governance Structure (Current Model)
See diagram: Governance Structure
3.1 Governing Team
The Governing Team provides strategic oversight and ensures:
- alignment with FactHarbor’s vision and mission
- compliance with legal and ethical obligations
- transparency of important decisions
- resolution of escalated issues that cannot be settled at domain level
The Governing Team focuses on *mission and integrity*, not micromanagement.
3.2 Executive Lead
The Executive Lead:
- coordinates the domains (Research & Development, Organisation, PR & Care & Marketing, Operations)
- ensures coherence and consistency across domains
- supports escalation handling and conflict resolution
- makes sure that governance rules are applied in practice
In a small organisation, the Executive Lead may also hold other roles, but responsibilities must remain clearly documented.
3.3 Governance Steward
The Governance Steward safeguards:
- neutrality of processes
- transparency of decisions
- fairness in conflict handling
- adherence to agreed governance rules
The Governance Steward is a focal point for concerns about process, fairness, or structural bias.
3.4 Advisory Roles
Advisors support decision quality without having direct decision authority:
- Legal Advisor – legal frameworks, contracts, licenses
- Ethics Advisor – ethical questions, conflicts of interest, societal impact
- Scientific / Domain Advisors – topic-specific expertise (e.g. medicine, energy, statistics)
Advisors may be consulted for specific questions; their input must be documented when it materially influences decisions.
3.5 Domain Leads
Each domain (R&D, Organisation, PR & Care & Marketing, Operations) may have a Lead who:
- owns day-to-day decisions within that domain’s boundaries
- escalates when decisions affect other domains or the whole organisation
- ensures that domain actions follow the agreed governance rules
4. Governance Model & Evolution (Future / Draft Path)
This section preserves important ideas from earlier governance drafts.
It describes a possible long-term path and does not override the current small-organisation reality.
Details may change before they are adopted in practice.
4.1 Stewardship Governance (Principle)
FactHarbor follows a stewardship governance approach:
- Strategic control remains with a trusted core to prevent hijacking or capture.
- Governance is designed to protect the mission rather than maximise profit.
- Power is exercised as a stewardship duty towards the public and contributors.
4.2 Startup Phase Governance (Founder-led)
In the early phase, governance was designed around a founder-led model, where:
- the Founder acts as a de-facto Sole Maintainer,
approving merges, managing releases, and holding final authority over technical and strategic decisions; - a Core Team may be added, with multi-party approval for security-sensitive or high-risk changes;
- a succession mechanism is expected to be defined before transition, for example:
- Founder-appointed successor, or
- successor ratified by a council-like body (e.g. future Governing/Steering council).
These ideas can be reused or adapted when the concrete legal and organisational structure is defined.
4.3 Possible Non-Profit Organisation Phase (e.g. Swiss Verein)
Earlier drafts envisioned a transition to a nonprofit entity (for example, a Swiss Verein) once FactHarbor reaches sufficient maturity and community scale.
Key ideas from that draft:
- Governance bodies might include:
- a Steering Council (central decision-making and strategic oversight),
- Core Maintainers (review and merge code / specification changes),
- a Security Council (security veto, audits, and sensitive decisions).
- The Founder’s role after transition could become:
- permanent or long-term member of the Steering Council,
- strategic vision holder, while decisions follow the agreed Charter.
- An Asset Transfer Protocol would be required when the nonprofit is formally created, e.g.:
- transferring copyrights, domains, repositories, and trademarks
- from the Founder (or initial holder) to the nonprofit entity
- in a well-documented, mission-locked way.
These points are preserved here as design material for future governance work.
They are not yet binding and must be confirmed, adapted, or replaced when the legal form is chosen.
5. Decision Processes
Decisions in FactHarbor are categorized and escalated according to specific protocols to ensure efficiency, fairness, and auditability.
For the full definition of decision types, escalation paths, and documentation requirements, see:
6. Compliance Framework
The Compliance Framework ensures that FactHarbor operates with legal adherence, financial transparency, and operational security.
For details on funding principles, ledgers, and internal controls, see:
The Governance page provides the high-level framework.
Details are further specified in the Organisation, Finance & Compliance, and Open Source Model & Licensing pages.
7. Core Design Goals
FactHarbor’s governance, open source model, and financing are built around a small set of long-term goals.
They collect ideas that are now spread across Governance, Open Source Model & Licensing, Finance & Compliance, and Legal Framework.
- G1 – Mission first, forever
The mission – clarity, transparency, and resistance to manipulation – must not be overridden by financial, political, or popularity incentives.
Governance and funding decisions are evaluated against this mission, not the other way round.
- G2 – Openness & Transparency
The reasoning engine, data processing, and the way AI support is used should remain inspectable and explainable.
The current licence mix (for code, documentation, and data) is chosen to: - keep core components openly usable and auditable, and
- make sure that any non-open pieces are clearly marked and governed.
For concrete licence choices, see Open Source Model and Licensing.
- G3 – Controlled Core, Open Contributions
Anyone may propose ideas and contributions, but FactHarbor relies on: - a small, trusted Governing Team and maintainer group for core decisions, and
- clearly documented contributor roles and processes.
This combination should keep the core coherent and safe, while still welcoming broad participation.
Details: Roles & Bodies, Contributor Processes.
- G4 – Financial Sustainability without Profit Extraction
FactHarbor aims to be financially sustainable without becoming profit-driven.
In practice this means: - revenue (donations, grants, services) is reinvested into the mission,
- no profit is distributed,
- key contributors can receive fair, market-aligned salaries when funding allows and law permits.
Details: Finance & Compliance.
- G5 – Manipulation Resistance
Governance and technical rules must: - prevent capture by hostile actors,
- protect against coordinated manipulation, and
- safeguard the integrity of claims, scenarios, evidence, and verdicts.
This affects both organisational structures (who can decide what) and technical design (audit trails, moderation tools, anomaly detection).
- G6 – Legal Clarity
Open source, governance, and financing must be: - legally defensible,
- compatible with relevant jurisdictions (e.g. Swiss, EU, US), and
- understandable for non-lawyers who need to work with the rules.
Details: Legal Framework.
These goals do not override more detailed rules on the subpages; they summarise the direction that Governance, Licensing, Finance & Compliance, and Legal Framework should remain aligned with.
8. AI, Transparency and Integrity (AKEL)
Because FactHarbor deals with truth-adjacent reasoning, any use of AI must meet higher transparency and integrity requirements.
- The AI Knowledge Extraction Layer (AKEL) is treated as part of the open core design.
Its purpose is to assist humans in extracting, organising, and updating knowledge – not to replace human judgement. - Where possible, AKEL should rely on open models or models whose behaviour can be reasonably inspected and documented.
- When proprietary or external AI services are used:
- this must be clearly disclosed to users at the point of use (e.g. in UI hints or context help),
- the system indicates why this model or service is used, and
- the core logic (how outputs are integrated, evaluated, and stored) remains open and auditable.
- AI outputs are treated as proposals, not as final verdicts.
Human review and governance rules decide what becomes part of the official knowledge base.
Licensing details related to AKEL and the core protocol are described in
Open Source Model and Licensing,
and the technical design is specified in the main Specification.
9. Evidence Openness
FactHarbor’s mission depends on open evidence practices. The core rules are:
- No hidden evidence
Evidence used in published reasoning should be accessible, or the limitations clearly documented (for example when data is confidential or privacy-relevant).
- No silent corrections
If a published statement is corrected, there must be a visible note or changelog entry explaining what changed and why.
- Versioned and traceable
Evidence collections, datasets, and key reasoning artefacts should be versioned.
It should be possible to reconstruct “what the project believed at time X”.
- Independence and conflicts of interest
Potential conflicts (for example funding, affiliations, roles) should be documented so users can judge possible biases.