State Liner · Car 03 · Deep Dive
Full doctrine treatment. Ten sections. Every claim with its reasoning. Every boundary with its hand-off.
This is the deep dive. The quick-start ride lives at /state-liner/03/. Same rail, one doctrine, two depths. Read whichever matches the time you have.
This is the long version of Car 03 — Section 0 through Section 10, unabridged and unedited. It reads as one continuous ride: the rail above tracks where you are, moves you section to section, and never loses your place. Rex narrates the Car itself; this longer route is the full reading, so it is labelled honestly as a reading rather than an audio ride.
You are the owner of a 25–100 employee company. You have thirty minutes to an hour. You want to leave with a signed one-pager, a review rhythm, and no obligation to buy anything from anyone including Dynasty.
This deep dive walks the same ten questions the quick-start walks, but with the reasoning left in. Every claim comes with its "why," and every boundary comes with its hand-off. If a paragraph feels obvious, skip it. If a paragraph feels uncomfortable, that is where the pitchmen are counting on you to bounce. Stay another minute.
The Only-Don paragraphs are marked. They are the owner's signed testimony and cannot be written by anyone else.
The AI governance market is a multi-billion-dollar industry that does not want the owner of a 25–100 employee company to have a clean answer to four questions. That is not a conspiracy. It is a business model. Every asset in the market — the $9,000 Trust Center, the $6,000-a-month vCISO, the $75,000 AI Readiness Assessment, the carrier surcharge questionnaire, the 40-page policy template — has the same structure: it makes money when you cannot answer the four questions on your own.
The four questions are:
Answer those four sentences in plain English on one page. Sign it. Review it for fifteen minutes once a month. That is the entire doctrine. Every product in the market above is priced against the assumption you will not do this. The moment you do, the assumption breaks.
This Car is designed to make you do it in the next thirty minutes.
Group AI governance in 2026 looks the way group health plans looked in the late 1990s: everyone agrees it is broken, no one wants to be the first person to name the specific players who profit from the brokenness. The fax-machine analogy holds. The fax machine was obviously obsolete for a decade before offices retired it, because three groups profited from its persistence: the toner sellers, the maintenance contract holders, and the compliance officers who insisted on faxable signatures. AI governance today has three equivalent groups.
Governance-as-a-subscription vendors charge more when you look like you might churn, less when you look like you are locked in, and always more when a new regulation drops. The subscription is priced against your fear of the next audit, not against the cost of running the thing. If the subscription solved the problem, it would end. It does not end because it is not designed to.
Every commercial carrier now has an AI section on the renewal questionnaire. The questions are structured so that "I have not decided yet" is priced as high risk, and "we have a signed AI governance framework" is priced as lower risk — even when the framework is a template downloaded from the vendor above. The carrier is not pricing risk. It is pricing ambiguity. That is why every carrier questionnaire rewards the owner who has signed something, regardless of whether the something is any good.
The $75,000 AI Readiness Assessment is the compliance industry's version of a diagnostic MRI: expensive, inconclusive, and always followed by a recommendation for more scans. The consultant invoices for the assessment, then the $6,000-a-month vCISO retainer to "operationalize the assessment," then the training, then the board packet, then the annual audit letter. Four line items. One doctrine you could have written yourself in fifteen minutes.
None of these three groups will tell you the four decisions are simple. Simplicity is the specific thing that ruins their bundle.
This Car is not written for the enterprise CIO. It is not written for the compliance officer at a Fortune 500. It is not written for the fractional CISO who is going to charge his own $6,000 a month. It is written for the owner-operator of a 25–100 employee company — the person who signs the payroll runs, reads the D&O insurance questionnaire on a Sunday afternoon, and gets asked by customers what the AI policy is without a warning.
You have six specific Monday moments where AI governance actually shows up in your business, and they are the reason this Car exists.
A customer emails or calls and asks whether their data ever touches AI in your workflow. The wrong answer is "let me get back to you." The right answer is "here are our four decisions, signed, and this is what they say about your data class specifically." One sentence, calmly delivered.
Your D&O, cyber, or professional liability renewal arrives with an AI section. The questions are designed to trip up an owner who has not signed a framework. If you have signed the four decisions, you fill the section out in five minutes. If you have not, you spend two weekends and still fill it out wrong.
An employee starts using a new AI tool without asking. Maybe they upload customer data into it. Maybe they generate work product with it and ship it. Without the four decisions signed, this event is a fire. With the four decisions signed, this event is a coaching moment against a policy the owner already wrote.
A board member asks what your AI policy is at the quarterly review. Answering "we are working on it" costs credibility that is measured in enterprise value at the next diligence event.
A new AI vendor pitches you a product with data-handling terms buried on page nine of the MSA. Without the four decisions signed, you cannot decide whether to accept the terms. With them, the four decisions read the MSA for you in three minutes.
Your team ships a piece of work an AI drafted. It goes to a customer. Something in it is wrong. Nobody can remember who reviewed it, because "reviewed" was never defined in your company. That gap is a doctrine gap, not a training gap.
Six Monday moments. Four decisions. One page. Fifteen minutes a month. If you can answer the six Monday moments cleanly, the rest of the AI governance market is trying to sell you something you do not need.
You do not need dollar figures on this Car to see the cost. If you cannot answer the four decisions on one page today, you are already paying for the absence in four categories every month.
Your team members are using AI tools you are paying for twice (same person on two seats), or tools your company does not know it is paying for (personal cards expensed), or tools your company is not paying for at all but should be (free tiers that touch customer data). None of these categories are captured on any single line item on the P&L. All of them are real spend.
AI-drafted work gets shipped, breaks, and gets fixed by a human at billable-hour rates because no receipt existed to catch the error at draft time. This is the single largest AI cost most 25–100 employee companies incur, and it is invisible because it hides inside "revision time" on existing project budgets. The absence of a receipt discipline is what makes it invisible. Signing the four decisions makes it visible. Making it visible is what lets you reduce it.
Your renewals cost more because the AI section is answered ambiguously. Your carrier prices the ambiguity, not the risk. The premium delta between "we have signed governance" and "we do not" is often larger than the cost of the assessment the pitchmen want to sell you. The four decisions on one page satisfy the carrier's underlying question without the assessment.
When a future acquirer, lender, or strategic partner runs diligence on your company, they will ask for twelve months of AI receipts. If you cannot produce them, the buyer discounts your enterprise value by an EBITDA multiplier. That is a real, measurable cost at exit — even though you are not selling this quarter. Every month without signed decisions and running receipts is a month of compounding diligence deduction.
The absence is not free. It is priced into every one of those four categories, every month you do not sign.
Dynasty maintains an internal doctrine with specific dollar ranges for each of these four categories, calibrated against 25–100 employee shops we have reviewed. The public Car will name those ranges only when they are externally source-backed. For now, treat the categories qualitatively — you will recognize each one when you look for it in your own P&L.
I am the person who decides how AI is used inside this company, and I am willing to name the four decisions on a single page, sign it, and review it for fifteen minutes once a month.
That sentence is not a policy. It is a stance. Every product in the AI governance market is priced against a different stance — some version of "I will pay someone else to decide this for me." The pitchmen have every incentive to keep the owner in that stance. This Car is designed to move the owner out of it in the next twenty minutes.
Named roles, named exceptions. Not "our team." Not "employees as appropriate." Specific roles allowed, specific roles restricted, specific exceptions granted by name and signed by you. When a new hire starts, the doctrine tells you what they are and are not allowed to do with AI on day one. When a role changes, the doctrine gets updated.
Named data classes, named prohibitions. Not "appropriate data." Named classes: customer PII, financial credentials, source code, board materials, employee health data, and so on. For each class, the doctrine says whether AI is allowed to touch it, under what conditions, and with what review. When a new data class appears in your business, the doctrine gets updated.
Named format, named retention. Every AI action that produces external work product leaves a receipt behind. The receipt names the person, the tool, the data class, the action, and the reviewer. Retention is at least twelve months so you can produce it for diligence, carrier questionnaires, or customer requests. The format is boring on purpose — a spreadsheet row is fine. What matters is that the receipt exists.
Named human, named cadence. For each category of AI-drafted external work product, the doctrine names the human who reviews it and the cadence on which review happens. Not "as needed." Not "the appropriate manager." A named person, a named rhythm.
Four decisions. One page. Signed. Reviewed fifteen minutes a month. The rest of the AI governance market is selling around this.
This Car is doctrine, not execution. When you have signed your four decisions and want help operating them, the rail bridges to the Dynasty stadium. The bridge is a doorway, not a sales pitch.
Rail fare: $28 per ride, comped through Sunday after Thanksgiving 2026. $28 per ride after that.
Why the fare exists at all: the pitchmen ship "free" e-books and then bill for the sales call. Dynasty's rail carries a transparent fare so the reader knows exactly what they are paying and what they are getting. Comping the fare through Thanksgiving weekend 2026 is a launch decision, not a permanent condition. Never call the rail "free" — the fare is the transparency.
The stadium has twelve named rooms; each hosts a master class inside it. Rooms are execution; the archetypes running the master class are the coaches. See the quick-start for the full room list.
Doctrine is what you can do in fifteen minutes a month on your own paper. That is what this Car gives you. Execution is what happens when you want tooling, board reporting, training, incident response, or a peer group of owners running the same discipline. Doctrine ships free of any sales pitch on this rail. Execution lives in the stadium rooms. Same rail carries you either way, at the same fare. Cross the bridge if execution is what you want. Do not cross it because I asked you to.
Ten commitments this Car makes against its own short-term revenue, so the doctrine stays a doctrine and never drifts.
Every future revision of this Car is checked against these ten lines before it ships. If a revision fails a line, it goes back for rewrite.
Ten paragraphs on this Car require the owner's signature and cannot be written by anyone else. This is what turns Car 03 from content into testimony.
Why an unsigned doctrine is a worse asset than a signed one: an unsigned governance doctrine is content. Content ages, gets dismissed, and gets priced against other content in the reader's inbox. A signed governance doctrine is testimony. Testimony gets stronger the longer it is correct and the more corrections it survives with a signed update log. Every vendor, carrier, and consultant in the market publishes unsigned content. Not one of them publishes signed testimony with a public correction log. This Car is the first widely circulated Dynasty asset that does.
Ten boundaries this Car names openly. Each one hands off to a specific Dynasty asset. A doctrine without stated boundaries drifts into every adjacent territory and loses its shape.
This Car names statutes; David M. Williams's GRC practice and the Federal Train Compliance Command Center read them against your specific situation. "You now have your four decisions on one page. Whether those decisions satisfy your specific statutory and contractual obligations is a Federal Train question."
Room 202 The Strategy Room diagnoses tooling for your specific company. Room 303 Training Camp builds it out. "Your four decisions dictate your receipt requirements. Your receipt requirements dictate your tooling. Which specific tools produce those receipts for your specific company is a Room 202 conversation."
Room 808 The Compliance Command Center handles incident response. "Car 03 is what you read on a Tuesday afternoon. If you are reading this at 2:47 AM because something broke, close this and open the Compliance Command Center. Different door, faster clock."
Room 606 The Dynasty Cast and Room 707 The EBITDA Engine produce board-ready reporting from twelve months of receipts. "Car 03 gives you the one-pager your board should have seen last quarter. What your board sees next quarter, backed by twelve months of receipts, is a Room 606 and Room 707 build."
Room 303 Training Camp and the Meridian Library architecture handle team training. "Your team learning how to live inside the four decisions is a Training Camp build, not a rail read."
This Car is a document, not a police force. Only the owner can enforce the doctrine, by reviewing receipts and holding people to the standard. "Car 03 gives you the standard. Whether your company meets the standard next Tuesday is a leadership question, not a document question. The document cannot fix that. You can."
SOC 2, ISO 27001, HITRUST, and sector-specific frameworks handle attestation; David's GRC practice runs the audit conversation. "If your customers require attestation, they require SOC 2 or its cousins. Car 03 is what you do before, during, and after the audit — not the audit itself."
The Meridian benefits cars (the SB 2610 / ACA seven-car system) and the benefits lane inside the stadium handle group health plan questions. "How AI touches your group health plan is a benefits-lane question. Meridian has an entire seven-car system for that."
This Car ends with a blank one-pager and a pen. The four decisions are the owner's to make, in the owner's voice, on the owner's paper. "Dynasty will not fill that page in for you — that is the entire point."
This Car does not renew. Read once, sign once, review fifteen minutes a month. "If you want more than that, more lives inside the stadium."
Why the boundaries themselves are the moat: every competing offer hides its boundaries because the money is in the ambiguity. This Car names them because Dynasty's money is on the other side of the reader being able to see clearly. The reader who can see the ten boundaries on this Car can also see the ten boundaries a $75,000 assessment refuses to name. Once the reader can see that, the assessment loses — not because this Car out-competes it, but because clarity out-competes bundling every time an owner has fifteen minutes and a pen.
This Car is a seed doctrine. Its job is to author the reader for the next five Cars on this rail, the two adjacent rails, and the three downstream assets. Every hand-off is published openly on the rail with no gate.
What a rail is: not a marketing funnel. A rail is an owned sequence of doctrine assets where each asset lowers friction for the next asset, every hand-off is published openly with no gate, and the reader can see the whole map. The pitchmen cannot. That transparency is why the rail compounds — the reader trusts the map because the map is visible, signed, and correctable.
Mr. Nobody, Founder, Dynasty Plays.
Signed 2026-09-02 · v1
The four decisions are your call, not mine. Take the blank page, take the pen, and sign your own doctrine. If any part of this Car turns out to be wrong, I will sign the correction here, in the log, in my voice. Fifteen minutes a month. One page. Your signature. See you on the rail.
Questions about this ride
Optional and up to you. Email David, GRC Chair at Dynasty Stack, at david.m.williams@dynastyplays.com. Nothing is shared and no one contacts you unless you write first.
Take the next step
← Back to quick-start · State Liner · Meridian Rail
Meridian Rail · State Liner · Car 03 · Deep Dive · v1 · 2026-09-02