AI Governance Advisor · Human Signal

Dr. Tuboise Floyd

The Leader Who Signs · TAIMScore™ Certified Assessor · Ph.D., Adult Education · Washington, DC

You approved the AI strategy. You bought the tool. Now somebody has to decide which use cases are worth the risk, who owns the decision when the system gets it wrong, what oversight is required, and what evidence you keep.

That work is where I come in.

Everyone in this field writes about frameworks. The frameworks are public. The gap is mine.
Join me in the Circle → Take the Two-Minute Test Connect on LinkedIn

Get the research as it publishes

Dr. Tuboise Floyd, Founder of Human Signal

The Claim

Belief versus execution is the whole of AI governance.

I come at this from an unusual direction.

My doctorate is in adult education, which is not a teaching credential. The field studies how people actually behave at work, measured against what they say they believe. My 2010 dissertation looked at workforce educators whose stated philosophy did not match their practice. The finding was that belief rarely survives the moment of execution.

That turns out to describe most AI governance failures. The policy says protect the data. The environment makes it easy to paste it somewhere else. The attestation says the control is in place, and the day-to-day says otherwise.

Technology has no beliefs. The gap sits in how people work under pressure, which is what I have studied for most of my career. I was studying that gap in 2010, before AI governance existed as a field to write about.

What Is Missing

Your policy is probably fine. The layer underneath is not.

Most leaders I talk with already have an AI policy. What tends to be missing is the layer underneath, the part that decides whether governance reaches the AI decision or stops at the document.

Intake

Risk tiering

Validation checkpoints

Harm testing

Approval pathways

Escalation design

The evidence trail you would want if somebody asked for it

I help you build that layer, and I show you where it runs thin before a board, a carrier, a regulator, or a contracting officer asks how your AI is governed.

Who I Work With

The person whose signature carries the exposure.

The president, provost, CEO, CFO, General Counsel, or CISO at a university, a mid-market finance, legal, or healthcare organization, or a defense industrial base contractor.

If that is your seat, I am glad to talk.

The Two-Minute Governance Test

One sentence decides which failures reach leadership.

A 17-page instrument that finds the sentence in your AI policy deciding which failures never reach leadership, then walks you through the rewrite before an assessor, a regulator, or the failure itself finds it first. Working time: one afternoon.

Open the door → humansignal.io/two-minute-test

Frameworks

What I Build

Original frameworks operators can actually use. Built from the pattern of how organizations break, not from vendor documentation.

Analysis

The Trust Gap

Two levels of organizational AI governance failure. Structural absence. Structural insufficiency. Permitted is not the same as admissible.

Read the framework →

Diagnostic

GASP™

Governance As a Structural Problem. Most organizations do not have a governance problem because they lack the right software. They have one because they never built the right structure.

Read the diagnostic →

Thesis

The Workflow Thesis

Distinguishes AI model failure from governance structure failure. Most organizations will not fail because of a bad model. They fail because of the structure around it.

Read the framework →

Practice

Noise Discipline

Cognitive defense for operators drowning in vendor hype. A structured practice for cutting through artificial noise and protecting organizational judgment.

Learn the practice →

Framework

The L.E.A.C. Protocol™

AI infrastructure viability across four physical constraints: Lithography, Energy, Arbitrage, Cooling. If your strategy does not address all four, you are leaking value.

Read the protocol →

Architecture

PSA® · AIaPI™

Presence Signaling Architecture and AI as Presence Interface. Frameworks for restoring human visibility in systems designed to observe, not listen.

Read the architecture →

Also in the canon: Hyperprompt™. Context control for LLM-enabled professional workflows.

The Platform

What We Publish

The AI Governance Briefing

An Apple Top 100 podcast on organizational failure, regulatory strategy, and enterprise risk.

The Failure Files™

Twelve governance failures mapped to the NIST AI RMF. Autopsies operators can learn from before the failure becomes theirs.

The AI Governance Record

A quarterly newsletter for boards, CISOs, and organizational operators. No vendor capture. No fluff.

Visible Human Research

Peer-track work including The Pedagogy Problem in AI Governance (SSRN, 2026), introducing Presence Signaling Architecture® (PSA®).

Signal & Silence

An audio novel on AI surveillance, algorithmic compliance, and organizational erasure.

Engagements

The practice runs as a ladder.

Each rung answers the question the last one raised. Start where the answer demands.

01 · Scan

The Two-Minute Governance Test

Scans an AI policy for the ten gaps a board, carrier, or auditor spots first, and scores how many are present.

Take the Test →

02 · Interpret

The Score Debrief

Turns the result into priorities. Thirty minutes on what your score actually means for your seat.

Book the Debrief →

03 · Defend

The Full Engagement

A baseline scored against the Trusted AI Model, 72 controls built by HISPI Project Cerebellum, conducted by a TAIMScore™ Certified Assessor.

See Advisory →

Organizational

Corporate Underwriting

Responsible AI founders and enterprise risk leaders fund the independent research layer, never the findings, and reach the operators who read the Record and listen to the Briefing.

See Underwriting →

Track Record

Experience

A career split between fixing systems under pressure and studying why they break.

Founder & Principal, Human Signal

AI governance advisory for regulated industries. The work serves the person who signs. The board packet, the carrier renewal, and the procurement attestation arrive on their own schedule, and the leader who certifies needs a defensible answer before they do.

Co-Lead, Project Cerebellum Advocacy & Education, HISPI

Co-lead the advocacy and education arm of HISPI's Project Cerebellum AI Think Tank under Col. Kathy Swacina, advancing responsible AI adoption and governance risk education across regulated industries. Elevated from Member to Co-Lead within two months.

TAIMScore™ Certified Assessor, HISPI

Credentialed to score an organization against the Trusted AI Model, the 72-control set HISPI Project Cerebellum built and maintains.

Board Director, Strategy & Modernization, APMP

Association governance under digital transformation pressure. An early case in board-level oversight of vendor selection and platform modernization.

Strategic Solutions Lead, Georgetown University

Led the launch of Hoyas Rising, Georgetown's Name, Image and Likeness collective, directing a cross-functional team of senior leaders through the NCAA legal and regulatory landscape. Ran the BDOG project portfolio across Workday implementations and a summer operations turnaround from $4.5M to $6M.

Education · Auburn University

PhD, Adult Education (Systems Theory) · BS, Business Management.

Independence

The yardstick does not move to fit my read.

The scoring runs against TAIM, a control set I did not write. HISPI holds the model. I carry the assessor credential, not the framework, and I sell no AI product of my own.

My lens sets the work. An external, NIST-aligned control set produces the number, so the answer holds up to the people who ask for it. Underwriters fund the research layer, never the findings.

Individual

Direct Contribution

A one-time or recurring contribution directly to the research. Every amount sustains the independence of this analysis.

Contribute →

Organizational

Corporate Underwriting

Organizations that share Human Signal's commitment to responsible AI can partner as named underwriters, with full editorial independence preserved.

See Tiers →

Grants

Research Grants

Foundations and public interest organizations seeking to support independent AI governance research are encouraged to reach out directly.

Get in Touch →

A Personal Invitation · The Human Signal Forum on Circle

Join me in the Circle.

Everything on this page. The frameworks, the Failure Files, the diagnostics. Gets worked live inside one room. Not a broadcast. A working room of operators who carry the same organizational weight you carry: the federal leads, the university executives, the CISOs, the board directors who have to make governance decisions inside systems they did not design.

I built the Human Signal Forum because governance is not learned from documentation. Governance is learned from failure, ideally someone else's, before it becomes yours. Full case treatment lands in the Reading Room before it publishes, and the Commons stays open for the file on your desk. No sponsor. No algorithm deciding what you see.

If you have been reading the Record, listening to the Briefing, or quietly running my frameworks against your own organization. This is the door. Come sit in the room where the work actually happens.

Take a seat. I will be in there.

. Dr. Tuboise Floyd

Join the Forum → Take a free Commons seat