◎ OS PUB Apache 2.0 ← All specifications

P268 — AIEP — Civilisation-Scale Bounded Reasoning Framework

Publication Date: 2026-04-12 Status: Open Source Prior Art Disclosure Licence: Apache License 2.0 Author/Organisation: Phatfella Ltd Schema: AIEP_OS_SPEC_TEMPLATE v1.0.1 — https://aiep.dev/schemas/aiep-os-spec-template/v1.0.1

Field of the Invention

[0001] The disclosure relates to bounded reasoning frameworks governing AI behaviour at civilisation-scale inference and decision-making contexts.

[0002] More particularly, the disclosure concerns a civilisation-scale bounded reasoning framework for use within AIEP federation networks, defining how AI systems operating at the scale of federated civilisation-level inference — affecting policy, resource allocation, scientific consensus, or large-scale societal decisions — apply mandatory epistemic bounds, uncertainty quantification, and human oversight obligations before producing outputs.

Framework Context

[0003] This invention operates within an Architected Instruction and Evidence Protocol (AIEP) environment as defined in United Kingdom patent application number GB2519711.2, filed 20 November 2025, the entire contents of which are incorporated herein by reference.

Summary of the Disclosure

[0004] A computer-implemented civilisation-scale bounded reasoning framework comprises: defining a CivilisationScaleThreshold as a configurable set of scope classifications triggering mandatory bounded reasoning, including: POPULATION_AFFECTED ≥ N, RESOURCE_ALLOCATION_VALUE ≥ V, SCIENTIFIC_CONSENSUS_DOMAIN, and CRITICAL_INFRASTRUCTURE_DEPENDENCY; applying a BoundednessCheckPipeline before any reasoning output is released at civilisation scale, comprising: (i) EpistemicBoundsCheck — quantifying uncertainty across all contributing evidence sets and rejecting outputs where aggregate uncertainty score exceeds uncertainty_ceiling without explicit GovernanceUncertaintyWaiver; (ii) OutcomeReversibilityCheck — requiring identification of reversal mechanisms for proposed actions and blocking outputs proposing irreversible civilisation-scale changes without multi-party governance sign-off; (iii) HumanOversightTrigger — mandating injection of a human review gate for all outputs passing CivilisationScaleThreshold classification, generating a HumanOversightRequest record before output release; (iv) CausualConsequenceProjection — running a minimum three-hop causal consequence projection over the Causal World State Graph (P200) and logging all identified second and third-order effects; generating a CivilisationScaleReasoningRecord as a first-class evidence artefact for all bounded reasoning events, capturing all check results, uncertainty scores, consequence projections, and human oversight decisions; and preventing automated action execution at civilisation scale without a signed HumanOversightApprovalRecord.

[0005] The technical effect is modification of computing system behaviour by mandating epistemic humility, consequence projection, and human oversight gates for AI outputs operating at scales affecting large populations, resources, or critical systems.

Claims

[0007] A computer-implemented method for civilisation-scale bounded reasoning comprising: classifying outputs against CivilisationScaleThresholds; applying EpistemicBoundsChecks with uncertainty quantification; requiring OutcomeReversibilityChecks; triggering HumanOversightRequests; executing causal consequence projections; generating CivilisationScaleReasoningRecords; and blocking automated civilisation-scale action without HumanOversightApprovalRecords.

[0008] A system for civilisation-scale bounded reasoning comprising one or more processors and a non-transitory computer-readable medium storing instructions to execute the method of claim 1.