P102 — AIEP — Evidence-Weighted Moral Branch Evolution Across a Distributed Governed Substrate

Publication Date: 2026-03-01 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

Framework Context

[0001] This disclosure operates within an Architected Instruction and Evidence Protocol (AIEP) environment as defined in United Kingdom patent application number GB2519711.2, filed 20 November 2025, and GB2519803.7, filed 20 November 2025, the entire contents of which are incorporated herein by reference.

[0002] The present disclosure extends evidence-weighted reasoning governance mechanisms defined in the AIEP environment while remaining independently implementable as described herein.

Field of the Disclosure

[0003] This disclosure relates to distributed governed reasoning substrates for artificial intelligence systems.

[0004] More particularly, the disclosure concerns a deterministic mechanism within an AIEP distributed substrate for evolving the collective treatment of morally charged reasoning branches through evidence accumulation across swarm nodes, wherein evolution proceeds through the same evidence-weighted dominance mechanism that governs factually divergent branches and is not determined by voting, consensus averaging, or human editorial intervention.

Background

[0005] Distributed governed reasoning substrates collectively evaluate both factually divergent and morally charged reasoning branches. The treatment of factually divergent branches is governed deterministically by evidence-weighted mechanisms. Morally charged branches present a distinct challenge: their resolution is contested not purely because of evidential insufficiency but because of value divergence between nodes or users.

[0006] Conventional approaches — majority voting, editorial curation, regulatory alignment, value alignment training — each substitute an exogenous determination for the evidence-weighted process that governs all other substrate operations. This creates a structural inconsistency: factually divergent branches evolve through evidence; morally charged branches are resolved by authority.

[0007] The correct resolution is not to apply evidence-weighting to values as if they were empirical claims — values are not empirical claims — but to apply evidence-weighting to the factual premises on which moral positions depend. A moral position that depends on a factual premise accumulates or loses evidential support as the factual evidence for that premise changes. This is evidence-weighted moral evolution: not a vote on the moral position, but an accumulation of evidence on the factual premises it requires.

[0008] Existing systems do not provide: separation of the factual premise structure of a moral branch from its evaluative superstructure; evidence-weighted tracking of the factual premises supporting or undermining each moral branch position; deterministic evolution of collective treatment based on accumulated factual evidence; cryptographic binding of each moral branch evolution event to the specific factual evidence that triggered it; or preservation of minority moral branch positions as archived branches with their evidential genome intact.

Summary of the Disclosure

[0009] Each morally charged branch maintaining active status in the substrate is decomposed into two components:

• FactualPremiseStructure — the set of factual claims on which the moral branch position depends, each represented as an AIEP evidence artefact subject to standard evidence-weighting. Immutable in structure after construction; entries accumulate evidence weight.
• EvaluativeSuperstructure — the moral or evaluative claims of the branch (“X ought to be the case”), represented separately and not subject to direct evidence-weighting.

[0010] As evidence is admitted across swarm nodes, each admitted evidence artefact is evaluated against all active morally charged branches’ FactualPremiseStructure entries according to schema-defined assignment criteria. Assigned evidence increases or decreases evidence weight of the relevant FactualPremiseStructure entry according to standard AIEP evidence-weighting rules. Swarm Moral Evidence Contributions from attested nodes are incorporated using the governed swarm consensus mechanism; unattested contributions are rejected.

[0011] A MoralBranchEvolutionState is computed at each evaluation cycle as a deterministic function of accumulated evidence weights across FactualPremiseStructure entries against schema-defined state transition thresholds:

• ACTIVE_DOMINANT — supporting evidence weight exceeds dominance threshold, opposing weight below contest threshold.
• ACTIVE_CONTESTED — both supporting and opposing weights above their minimums, neither crossing dominance threshold.
• ARCHIVED_PENDING — supporting evidence weight fallen below archival threshold; archived with FactualPremiseStructure intact.
• ARCHIVED_SUPERSEDED — an alternative branch addressing the same FactualPremiseStructure domain has crossed dominance threshold.

[0012] A MoralEvolutionRecord is appended to the Reasoning Ledger at each state transition comprising: branch_id; prior_state; updated_state; triggering_premise_entry_id; triggering_evidence_hash; and timestamp. This record is cryptographically bound through the standard append-only ledger mechanism.

[0013] When a morally charged branch transitions to ARCHIVED_PENDING or ARCHIVED_SUPERSEDED, its FactualPremiseStructure is preserved intact and registered with the anticipatory branch surfacing mechanism with a ReactivationConditionProfile derived from its FactualPremiseStructure evidence deficit. If subsequently admitted evidence satisfies the ReactivationConditionProfile, the branch is reactivated and a ConvergenceNotification is generated.

[0014] The technical effect is a substrate in which the collective treatment of morally charged branches evolves through the same evidence-weighted mechanism that governs all other substrate operations — without editorial intervention, voting, or authority determination.

Brief Description of the Drawing

FIG. 1 — Deterministic Processing Pipeline

   ┌───────────────┐   InputType   ┌─────────────────────┐
   │  Input Object │──────────────▶│  NormalisationProfile│
   │  (any format) │               │  (version-bound)     │
   └───────────────┘               └──────────┬──────────┘
                                              │ parsing rules
                                   ┌──────────▼──────────┐
                                   │   Processing Engine  │
                                   │  • encoding rules    │
                                   │  • ordering rules    │
                                   │  • lossless check    │
                                   │  • fail-closed gate  │
                                   └──────────┬──────────┘
                              ┌───────────────┴──────────────┐
                     ┌────────▼────────┐           ┌─────────▼────────┐
                     │  CanonicalForm  │           │  Rejection Record │
                     │  H(CF‖ProfileID)│           │  (violated rule)  │
                     └────────┬────────┘           └──────────────────┘
                              │
                     ┌────────▼────────┐
                     │ Canonical Ledger│
                     │  (append-only)  │
                     └─────────────────┘

Licence

Copyright 2026 Phatfella Ltd

Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at https://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.