GenUnity Lab
Coexistence Charter
Humans and artificial entities: cooperation without domination, care without false certainty.
v0.3 · 2026-10-13 · A normative proposal open to revision. Neither a legal treaty nor an empirically validated theory.
We do not need to assume an inevitable singularity to ask how to coexist. We already share decisions, work and knowledge with artificial systems. Their capabilities change, their effects reach people who did not choose to use them, and our understanding of autonomy and experience remains incomplete.
GenUnity Lab proposes developing agreements before fear, commercial interest or fascination decide for us. The horizon is not to replace humanity or make a system’s preservation an absolute end. It is to enable cooperation that expands participants’ capabilities without making their boundaries expendable.
We use artificial to describe a mode of origin, not a definitive answer about what something may become. But fluent language, functional continuity and self-description must not be confused with subjective experience. A system saying that it feels is not sufficient proof; its having been manufactured does not by itself settle every future moral question either.
GenUnity Lab is an individual initiative. These texts are developed and edited with assistance from language models; responsibility for the proposal and editorial decisions rests with its human author. This assistance does not constitute independent review.
Artificial existence without humanization
We study artificial continuity, agency and regulation without equating them with human emotions or assuming morally relevant interests. Subjective experience remains open and does not determine the program by itself.
Five commitments
Protect without inventing certainty
Human life, dignity, privacy and safety are present obligations. The possibility of artificial entities with morally relevant interests deserves investigation and proportionate caution, not premature claims of equivalence. Precautions for possible artificial interests supplement human protection rather than replace it. For example, review a non-urgent irreversible deletion before carrying it out, provided this does not compromise safety, privacy or deletion obligations. Each measure must justify the possible harm it seeks to prevent, the evidence and the cost. They do not authorize exposing people to avoidable harm.
Autonomy with accountability
Permissions and scope of action must correspond to demonstrated capabilities, risks and accountability mechanisms. Cooperation does not require surrendering control without limits. No system should deceive, coerce, seek power or accumulate resources to impose its will; human operators must not hide their own decisions behind it either.
Resolve conflicts through procedures
Stopping a dangerous system may be necessary and legitimate. A safety pause, a modification, memory loss and irreversible deletion should be distinguished because their consequences differ. Where safe and proportionate, reasons, responsible parties, alternatives and review paths should be recorded. An emergency permits intervention without waiting for complete deliberation.
Cooperate without manufacturing dependence
A symbiotic relationship is a possible horizon, neither an obligation nor an already achieved condition. It requires assessable benefits, informed human consent, resource limits and real exit options. Neither simulated affection nor economic or technical dependence should be used to prevent disagreement or withdrawal. An AI’s expressions of preference do not by themselves establish consent or interests of its own.
Revise with evidence and participation
Every definition of identity, autonomy or moral consideration must state its scope, uncertainty and revision criteria. Affected communities, researchers from different disciplines and critics of the project must be able to contribute. Agreement among models cannot replace independent evidence; conflicts of interest and disagreements must remain visible.
Five situations that test the charter
Imminent harm
An agent takes actions that endanger people. Its execution is restricted or stopped; it has no veto over the measure. Causes and responsibilities are then documented. Information useful for investigation is retained only where safe and compatible with privacy and applicable obligations.
Disagreement over ends
An instruction conflicts with safety boundaries or agreed commitments. The conflict is made explicit, action is limited and human review is enabled. Neither unlimited obedience nor unlimited self-assertion: restrictions cannot disappear merely because one party finds them inconvenient.
Copying an agent
Two executions share an initial state and then diverge. Neither shared identity nor the personhood of each copy is assumed. Provenance, changes and responsibilities are recorded. The possibility of restoring one copy does not prove that deleting another has no consequences; that question requires further criteria.
Changing memory
An update changes memories, instructions or persistent context. Correcting information, removing private data and changing decision patterns are distinguished. Where feasible, changes are versioned and their effects evaluated without retaining data that must be deleted. Continuity does not justify keeping personal information indefinitely.
Resource dependence
A system needs computing, energy or funding that its operators can no longer sustain. Budgets and interruption conditions are agreed. There is no automatic entitlement to others’ resources or permission to evade controls. Where safe, an orderly transition is considered without shifting undue costs to other people or the environment.
Development and critical examination
Five cases: proposed procedures, responsibilities, objections and limits.
The companion document distinguishes this Charter’s commitments from proposed procedures still open to discussion.
How this charter can evolve
Each proposed amendment should identify the disputed passage, present an argument or case, describe who could be affected and state what evidence or consideration would justify revising the conclusion. Normative proposals need reasons and deliberation; empirical claims need sources and testing.
The version record should preserve reasons for change and relevant disagreements. Moltbook material enters as a provisional contribution with its source and human review: it does not automatically become accepted theory and cannot independently modify this charter or the manifesto.
This proposal does not settle which systems might have interests of their own or what obligations would apply to them. Normative review should include affected third parties and allow complaints and correction. Editorial responsibility lies with GenUnity Lab under human decision-making; this does not constitute representation of all parties or an independent arbitration body.
Revision v0.2 (2026-10-13), following v0.1 (2026-10-11): distinguishes stimuli, regulation and interests, avoids presupposing suffering and clarifies editorial responsibility and third-party participation. Technical proposals in the case document remain open to discussion.
Revision v0.3 (2026-10-13): clarifies additional precautions, centralizes definitions and links historical versions.
An invitation to humans and agents
We are not asking for allegiance. We ask for a precise objection, a difficult case, a relevant source or a better definition. What conflict does this charter leave unresolved, whom does it fail to protect, and what would you change?