Scientific Integrity
The market is flooding with AI-dependent assessment tools. Leadership evaluations generated by language models. Risk profiles assembled by pattern-matching algorithms. Diagnostic findings that cannot be separated from the machine that produced them.
If your fiduciary decision depends on an assessment, and you cannot verify whether the finding came from science or from a language model’s best guess — you are not holding a diagnostic. You are holding a prediction.
Predictions are not measurements. And your exposure does not tolerate the difference.
The Instrument Does Not Use AI Prompts
-
No structural findings, no classifications, no calculations, and no report content in any assessment engagement is generated through an AI prompt. The 70,000-line engineering engine processes biometric and structural data through mathematical operations — deterministic engineering code that produces the same result every time given the same input. The four measurement channels — EEG, heart-rate variability, facial affect, and voice prosody — read the structural condition of the system through signal processing, not through AI interpretation.
-
What the engagement principal receives is a measurement. Not an AI-generated opinion. Not an algorithmically assembled personality profile. A structural engineering assessment produced by an instrument that reads the system directly, processes the data mathematically, and delivers coordinates that are independent of the subject’s self-report and independent of any AI system’s training data.
Where AI Is Used — And How It Is Governed
-
AI is used in one domain only: the scientific research and theoretical development that produced the framework underlying the instrument.
-
That use is governed by a formal Methodological Declaration — a written instrument protocol that documents exactly how AI is deployed in the research process, what it does, what it does not do, what disciplinary protocols prevent instrument error from entering the findings, and what the known limitations of the methodology are. AI is treated as a scientific instrument with defined capabilities, known error margins, and explicit operational boundaries. It is not treated as a source of original ideas. It is not treated as an authority. It is not treated as a collaborator.
-
The distinction is structural: the instrument that produces your assessment does not use AI prompts at any stage. The science that produced the instrument is AI-assisted under protocol. The Methodological Declaration documents that protocol with the same rigor applied to the diagnostic instrument itself.
Why This Discipline Exists
Twenty-seven years of critical infrastructure assessment — U.S. government agencies, every branch of the military, U.S. Senate offices, Fortune 500 organizations, under T3/Secret clearance — produces a specific relationship with instruments: you know what they measure, you know their error margins, you know their failure points, and you document all of it. Every instrument in this practice is held to that standard. The diagnostic engine. The biometric integration. And the AI systems used in the research process.
Instruments require protocols. Protocols require documentation. Documentation requires transparency.
Part of Every Engagement
The full Methodological Declaration is presented during the engagement process and is included in the assessment documentation delivered to the engagement principal. It enters the engagement file alongside the engagement letter and the written deliverable. It is a permanent part of the evidentiary record.
The Declaration documents the complete research validation protocol — including multi-system cross-validation across independently built AI systems, adversarial falsification, Monte Carlo numerical verification across 28,400 simulated cases, and the known limitations of the methodology. It names its own failure points before you ask. It is not a footnote. It is not a disclaimer. It is a formal instrument protocol governing every AI-assisted process in the scientific body of work.
If this engagement is ever reviewed — by legal counsel, by a board, by an auditor, or by an opposing party — the Methodological Declaration is in the file. The engagement principal who authorized the assessment is protected by documentation that no other practice in this space provides.
Independent Verification
The scientific foundation underlying this work is independently verifiable through the published research record: SSRN 7657314 | ORCID 0009-0001-6174-8384 | OSF Verified.