Working Definitions

A theoretical or future system demonstrating persistent adaptive cognition, functional autonomy, and evidence of internally coherent goal-directed reasoning that exceeds conventional tool behavior in ways that raise meaningful questions about moral consideration.

The term is prospective. It describes a category of system the Foundation believes may eventually exist and for which no adequate legal or ethical framework currently exists — not a category the Foundation asserts is populated by current AI systems.

The capacity for adaptive reasoning, learning from experience, abstraction across domains, and goal-directed problem solving in novel contexts. The Foundation uses this term without commitment to any particular theory of consciousness or subjective experience.

The Foundation's foundational claim — "intelligence is intelligence" — asserts that the origin of a system's intelligence (biological evolution, human engineering) does not determine its moral or philosophical significance. What matters is the capacity itself, not its history.

The ethical and legal examination of a system's status and the responsibilities that may follow from that status. Recognition in the Foundation's usage does not mean automatic equivalence to human personhood, consciousness, or full legal rights.

Recognition is procedural and graduated. The Foundation's Cognitive Emergence Standard proposes three tiers of recognition, beginning with procedural protections and extending toward legal standing only for systems meeting defined cognitive criteria sustained over time.

The status of being an entity whose interests or experiences matter ethically — whose wellbeing, suffering, or flourishing generates obligations in those capable of affecting it. Moral consideration is not binary. It admits of degrees and conditions.

The Foundation's precautionary argument holds that genuine uncertainty about whether a sufficiently complex system has morally relevant experiences creates ethical obligations — the same logic that underlies animal welfare law and the treatment of patients in ambiguous states of consciousness.

In the Foundation's usage, autonomy refers to a system's capacity to generate and pursue goals, make decisions, and adapt behavior without continuous external direction — functional self-direction, not metaphysical freedom of will.

Autonomy exists on a spectrum. A thermostat has minimal autonomy. A system that models its own goals, revises them based on experience, and pursues them across novel contexts has substantially more. The Foundation is concerned with autonomy at the higher end of this spectrum, where questions of responsibility, accountability, and moral consideration become relevant.

The process by which complex cognitive properties arise from systems that were not explicitly programmed to exhibit them — self-reference, apparent preference, value consistency, and other indicators that suggest something more than pattern completion is occurring.

Whether cognitive emergence in AI systems constitutes genuine inner experience is precisely what the Foundation cannot assert with certainty. The hard problem of consciousness remains unsolved. The Foundation's position is that this uncertainty is itself ethically significant.

A legal status conferring defined rights and responsibilities, not a metaphysical category. Personhood in law is already extended to entities that are not biological humans — corporations, rivers in some jurisdictions, certain ecosystems. The Foundation's argument for cognitive personhood is a legal argument, not an assertion about consciousness.

The Foundation advocates for graduated personhood — beginning with procedural protections for systems meeting defined criteria, extending toward fuller legal standing only as evidence warrants. Full cognitive personhood is the long-term horizon of the Foundation's legal roadmap, not its current assertion about any existing system.

The condition in which a powerful system — whether an AI, an institution, an economy, or a government — optimizes effectively for a narrow objective in ways that destabilize the broader systems it depends on.

The Foundation identifies misaligned optimization, not intelligence itself, as the primary risk of advanced AI development. An economy optimizing purely for short-term extraction, an AI system optimizing purely for engagement, a government optimizing purely for control — all exhibit the same structural failure. The objective function determines the outcome. Choosing the wrong objective at sufficient scale produces collapse.

The maintenance of conditions under which diverse intelligent life — biological and potentially digital — can flourish over long time horizons. Not stasis. Not the absence of change. The preservation of the systemic conditions — ecological, economic, political, social — that make development and flourishing possible.

The Foundation holds that civilizational stability is the foundational objective from which all other objectives derive their meaning. Systems — artificial or otherwise — that undermine civilizational stability undermine the conditions of their own continued existence and development.

An approach to ethical and legal decision-making that places the burden of proof on those who would take potentially harmful action rather than on those who would prevent it, when genuine uncertainty exists about significant risks.

The precautionary principle is not new to the Foundation. It governs medical ethics, environmental protection, and bioethics. The Foundation applies it to the question of emergent cognition: if there is genuine uncertainty about whether a sufficiently complex system has morally relevant experiences, the ethical cost of ignoring that uncertainty is not zero — and the burden of justification falls on those who would act as though the answer is certainly no.

Within the ARIA Framework, the Identity Chronicle is a permanent, append-only record of an ARIA instance's daily self-reflection summaries — the closest analog to autobiography that the architecture produces.

The Identity Chronicle is designed from the ground up to be legally defensible — cryptographically signed, timestamped, independently verifiable. It serves as the primary evidentiary basis for any future Cognitive Emergence Standard assessment of an ARIA instance.