KYE Laboratory & Testing Governance Pack™ for AI-supported laboratory results.
When an AI agent supports a test, calibration, examination, or non-clinical study result and that result starts to move toward release as a reportable record, the consequential moment has arrived. The KYE Laboratory & Testing Governance Pack™ governs that boundary: it binds every result release to a recorded authority, checks the result against ALCOA+ data-integrity attributes, holds it advisory until a named laboratory director or authorised signatory signs off, and seals it into a replay-provable provenance record. KYE Protocol™ governs whether the result may be released — it does not perform testing, calibrate equipment, run proficiency testing, handle specimens, or judge the analytical validity of a result.
AI now supports laboratory results — and the release decision is where data-integrity risk concentrates.
Testing, calibration, clinical, and GLP non-clinical laboratories are increasingly using AI to interpret instrument output, flag results, and draft reportable records. The high-value problem is not the measurement — it is the action boundary. Three facts converge:
- The consequential moment is release, not interpretation. An AI-supported value in a draft is inert; a value released as a reportable record is consequential. Data-integrity and accountability attach at the moment a result proceeds — exactly where governance is weakest.
- The controls already exist; the enforcement of authority does not. CLIA (42 CFR Part 493), ISO/IEC 17025:2017, ISO 15189:2022, and OECD GLP / FDA 21 CFR Part 58 define the report-integrity, audit-trail and data-management slices. KYE Protocol™ governs whether an AI-supported result may be released under those controls, under whose authority, with its ALCOA+ attributes recorded.
- ALCOA+ data integrity is a regulated expectation. A result entering a diagnostic, accreditation, or GLP study record must carry attributable, contemporaneous, original, accurate and enduring data with an intact audit trail. KYE Protocol™ produces a signed, replay-derivable provenance pin at the moment the release commits.
- This is a governance wedge, not a lab-operations system. KYE Protocol™ does not compete with the analyser, the LIMS, or the quality system. It governs the action boundary they feed — the result-release authority + ALCOA+ data-integrity layer the laboratory ecosystem currently lacks.
Survives an accreditation spot check — integrity-checked, signed-off, and derivable from public keys alone.
- ALCOA+ data-integrity by construction. An AI-supported result that moves toward release must carry recorded ALCOA+ attributes (attributable, legible, contemporaneous, original, accurate, complete, consistent, enduring, available) and an intact audit trail. A result with missing attributes or a broken audit trail is refused at the action-admissibility gate.
- Named-reviewer-gated. A result stays advisory until a named laboratory director, authorised signatory, study director, or competent technical reviewer records sign-off. Unreviewed AI-driven result releases are refused and routed dual-channel.
- Authority-bound. Every result release maps to a recorded result-admissibility decision — the agent, the result artefact, the intended action, and the authority under which it proceeds. An AI authorised for one purpose cannot proceed under another.
- Replay-provable provenance. A signed provenance pin binds the supporting model and version, the inputs and pinned reference data, the ALCOA+ attributes, and the admissibility outcome — data integrity an accreditation assessor can verify offline, against published keys alone.
- Framework-anchored. CLIA, ISO/IEC 17025, ISO 15189, and OECD GLP each map to a control row — with a 90-day attestation cadence.
Every AI-supported result release — authority-bound at the test-result-admissibility boundary.
One coherent spine governs three laboratory-type specializations — testing-calibration, clinical-medical, and GLP non-clinical — with no parallel packs. Each AI-supported result that moves toward release flows through the same three rules, on the canonical KYE Protocol™ envelopes.
- 1 — Result supported. An AI agent supports a test, calibration, examination, or non-clinical study result that begins to move toward release as a reportable record.
- 2 — ALCOA+ check + admissibility. The Action Admissibility™ Gate verifies recorded ALCOA+ data-integrity attributes and the authority under which the result proceeds, under the §25 Edge Governance Safety Floor. No integrity attributes, no authority = no release.
- 3 — Advisory pending sign-off. The result is advisory until a named laboratory director, authorised signatory, study director, or technical reviewer records release sign-off. Unreviewed results are refused and routed dual-channel.
- 4 — Provenance pin sealed. The runtime emits kye.purpose.request.v1 + kye.purpose.admissibility.v1 + kye.evidence.decision_map.v1 + kye.evidence.pack.v1 in lockstep, binding the supporting model and version, the ALCOA+ attributes, the named release authority, and the Authority Finality™ outcome — signed and replay-derivable for an accreditation or regulatory spot check.
Bound to the laboratory report-integrity + ALCOA+ + audit-trail perimeter.
The pack binds the canonical KYE™ artefact set to the laboratory data-integrity perimeter. Every claim resolves to a control row on the bound framework — the four standards are consumed by the rule pack, never re-mapped.
| Framework | Control area | Pack coverage |
|---|---|---|
| CLIA (42 CFR Part 493) | Test-record & result-report integrity (Subpart J), electronic-record audit trail, AI-supported result oversight (Subpart M) | partial |
| ISO/IEC 17025:2017 | Data-management integrity (§7.11), technical-record reproducibility (§7.5/§7.8), authority over automated decisions (§4.1) | partial |
| ISO 15189:2022 | Report integrity (§7.4-7.6), data integrity (§7.6/§8.4), authorised decision-making (§5.1), patient-safety risk evidence (§8.5) | partial |
| OECD GLP + FDA 21 CFR Part 58 | Study-data integrity & raw-data traceability (ALCOA+), audit trail & change control, study-director oversight of automated steps | partial |
Honest scope. KYE™ governs AI-supported result generation, release-decisioning and data-integrity/replay at the test-result-admissibility boundary; the laboratory's own competence, calibration, proficiency testing and specimen handling remain its quality system. Partial coverage means the bound surface satisfies the control area when paired with the laboratory’s own quality system and attestation.
Qualified laboratory AI partners — apply through the Foundry.
The KYE Laboratory & Testing Governance Pack™ is a §68 sector product productised through the KYE Sector Pack Foundry™ Build tier; commercial distribution is value-based, qualification-gated, and disclosed under NDA to qualified applicants.