P140 — AIEP — Evidence Market Protocol

Publication Date: 2026-03-15 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 defines an open protocol for institutional and organisational participation in the AIEP evidence ecosystem, enabling entities to publish structured evidence records as signed EvidencePublications that are discoverable, trustable, and reusable by AIEP-governed reasoning systems.

Field of the Disclosure

[0003] This disclosure relates to open evidence market protocols for governed artificial intelligence ecosystems.

[0004] More particularly, the disclosure concerns a schema and publication procedure under which organisations become certified evidence providers, sign and publish structured EvidencePublication records, accumulate a cryptographically verifiable ProviderTrustScore over time, and specify usage licences that govern how their evidence may be consumed by AIEP-governed reasoning systems.

Background

[0005] AIEP-governed reasoning systems require evidence artefacts with verifiable provenance. The Distributed Evidence Index Protocol (P133) defines how evidence nodes are indexed and discovered, but does not define the governance framework under which organisations publish evidence, build reputation, or declare usage rights.

[0006] Without a normative evidence market protocol, consuming reasoning systems have no standard mechanism to verify the identity or reputation of an evidence publisher — only the cryptographic integrity of the content hash. A publisher’s historical accuracy, institutional credibility, and citation record are valuable signals for evidence weighting that require a structured reputation mechanism.

[0007] Existing knowledge publication approaches — academic preprint servers, regulatory document portals, open data registries — do not attach cryptographic identity to publishers, do not accumulate machine-readable reputation records, and do not declare usage rights in a format consumable by AI reasoning systems.

Summary of the Disclosure

[0008] An EvidenceProviderNode represents an organisation participating in the evidence market: provider_id derived as the SHA-256 hash of the organisation’s canonical name and registration timestamp; organisation_name; jurisdiction; domain_specialisation as a schema-defined domain taxonomy array; public_key as an ECDSA P-256 public key in JWK format; and provider_trust_score initialised to a schema-defined baseline.

[0009] Each EvidencePublication record comprises: evidence_id derived as the SHA-256 hash of provider_id, source_url, and publication_timestamp; provider_id; source_url; mirror_url; content_hash; publication_timestamp; classification_tags; jurisdiction; language; domain_category; and a base64-encoded ECDSA-P256 signature over the canonical serialisation of all fields excluding the signature. Consuming systems verify the signature against the provider’s published public key before admitting the publication.

[0010] ProviderTrustScore is computed as a deterministic function of: citation frequency (the count of reasoning chains referencing the provider’s publications); historical accuracy (the proportion of the provider’s publications that have not been contradicted by admitted negative proofs); and institutional credibility (a schema-defined tier derived from the provider’s public regulatory registration, if applicable). The ProviderTrustScore is updated after each evidence admission event that references the provider.

[0011] Evidence Licensing. Each EvidencePublication may carry an EvidenceLicence specifying: licence_type (open_public, restricted_domain, commercial); usage_restrictions as a free-text field; attribution_requirements; and jurisdictional_limitations. Consuming AIEP systems must evaluate EvidenceLicence terms before admitting a publication to the evidence rail.

[0012] Evidence Discovery. When a reasoning system searches the distributed evidence index for candidates, it may filter by provider_trust_score minimum, jurisdiction, domain_category, and recency. The ProviderTrustScore filter operates as a soft pre-ranking rather than a hard gate: publications from lower-trust providers may still be admitted if they pass the governance admissibility threshold.

[0013] Provider Certification. An optional certification layer — administered by the AIEP Foundation or a designated certification body — may issue a CertificationRecord to providers who meet defined standards of institutional credibility, accuracy history, and public key management. CertificationRecords are published at the provider’s well-known endpoint and are included in ProviderTrustScore computation.

Technical Effect

[0014] Cryptographic signing of EvidencePublications by provider ECDSA keys enables consuming systems to verify publisher identity and publication integrity in a single verification step, without trusting any intermediary.

[0015] ProviderTrustScore derived from citation frequency and accuracy history — updated deterministically after each relevant evidence admission event — provides a machine-readable, objective reputation signal that consuming systems can use for evidence pre-ranking without manual curation.

[0016] Open publication of the evidence market schema, signing procedure, and trust score computation as prior art prevents competitors from claiming proprietary rights over the institutional evidence publication layer, ensuring interoperability across the AIEP evidence ecosystem.

Claims

1. An evidence market protocol for a governed AI ecosystem, the protocol comprising: an EvidenceProviderNode schema including a deterministically derived provider identifier, an ECDSA P-256 public key in exportable format, jurisdiction, domain specialisation, and a provider trust score; an EvidencePublication record schema including a derived publication identifier, provider identifier, source URL, content hash, publication timestamp, classification metadata, and an ECDSA-P256 signature over the canonical serialisation of the record; and a signature verification procedure applied by consuming systems before admitting a publication to the evidence rail.

2. The protocol of claim 1, wherein the ProviderTrustScore is computed as a deterministic function of citation frequency in governance-admitted reasoning chains, historical accuracy as the proportion of publications not contradicted by admitted negative proofs, and an institutional credibility tier derived from public regulatory registration status.

3. The protocol of claim 1, wherein each EvidencePublication may carry an EvidenceLicence specifying licence type, usage restrictions, attribution requirements, and jurisdictional limitations, and wherein consuming AIEP systems evaluate EvidenceLicence terms as a pre-condition of evidence admission.

4. The protocol of claim 1, wherein evidence discovery may filter candidates by provider trust score minimum, jurisdiction, domain category, and recency, with the trust score filter operating as a soft pre-ranking of candidates rather than a hard admission gate.

5. The protocol of claim 1, wherein an optional CertificationRecord issued by a designated certification body is published at the provider’s well-known endpoint and incorporated into ProviderTrustScore computation, enabling tiered trust distinctions between certified and uncertified providers.

Brief Description of the Drawing

FIG. 1 — EvidencePublication signing and verification flow: provider key generation → publication construction → ECDSA signing → publication to index → consumer signature verification → admissibility gate.

FIG. 2 — ProviderTrustScore computation inputs: citation frequency, accuracy history, institutional credibility tier.

Abstract

An evidence market protocol enables organisations to become certified AIEP evidence providers by generating an ECDSA P-256 key pair, constructing EvidencePublication records with canonical metadata, signing records with the provider key, and publishing to the Distributed Evidence Index. Consuming systems verify provider signatures before admission. A ProviderTrustScore accumulates deterministically from citation frequency, accuracy history, and institutional credibility. EvidencePublications may carry an EvidenceLicence declaring usage restrictions. An optional CertificationRecord layer enables tiered provider trust distinctions. The full protocol schema is published as open prior art.

Licence

Apache License 2.0 — https://www.apache.org/licenses/LICENSE-2.0

Copyright 2026 Phatfella Ltd. Licensed under the Apache License, Version 2.0. You may use this specification in compliance with the Licence.