# Re-appraisal: CAIN-42 as an Agent Trust Fabric / Governance Control Plane

**Valuation date:** 14 October 2026 · **Prepared by:** Claude (AI assistant), at the owner's request
**Corrects the category** in `APPRAISAL_STYLE_VALUATION_2026-10-14.md`; **reconciles** it with the IP register.
**Standard of value:** fair market value · **Premise:** going concern, asset sale as floor

> **Read this first.** Not a certified appraisal (no ASA/ABV/CVA, not USPAP-compliant); not for tax, estate,
> lending, litigation or securities use. The preparer is not independent of the subject. Nothing here is a
> forecast, a financing valuation, or a guarantee. Paper equity, not cash.

## 1. Category correction (what actually changed)

Earlier framings drifted toward "AI security" language — a **scanner / security tool**. That was a **category
error** and it distorted the comparable set.

- **What CAIN-42 is:** an **agent trust fabric and governance control plane** for autonomous AI — the runtime
  that decides whether an agent's proposed action is authorized, enforces that decision *before* execution, and
  records signed, re-checkable evidence. Infrastructure, not a point security product.
- **Correct peer set:** agent control-plane / agent-identity / MCP-governance companies (Guild.ai, Beltic,
  Entro, Natoma, Zenity, Credo AI) — **not** AI-security scanners (HiddenLayer, Lakera, Protect AI).
- **Effect:** changes the comparable universe, positioning and opportunity ceiling. It does **not**, by itself,
  change the fundamentals that set fair market value today (verified revenue, audit, incorporation, assigned
  IP, patent, team). Positioning is a narrative asset until an arm's-length transaction prices it.

## 2. Two honest numbers (do not conflate them)

CAIN-42 carries **two different, clearly-labelled figures**. Both are legitimate; they measure different things.

| Basis | Figure | What it measures |
|---|---|---|
| **IP register scenario** (v55; 168 invention families) | **base ≈ $16.34M**, range **$6.34M–$28.90M** | bottom-up worth of the built mechanism/IP catalogue, valued at engineering/replacement cost. A **register scenario**, not what a buyer pays. |
| **Fair market value appraisal** (cost/market/income) | **≈ $0.47M**, range $0.26M–$1.0M | what an arm's-length buyer would pay for the asset **today**, given no verified revenue, no incorporation, no audit, no patent, one person. |

**The owner's range — $10M–$30M — is the IP register scenario band** (it sits inside $6.34M–$28.90M), and it is
also the **Stage-3 target** below. It is **not** the fair market value today, and the two must never be merged
into one headline.

## 3. Why the register is high and FMV is low

- The register sums what was **built** (168 families, each costed) — a supply-side, catalogue number.
- FMV asks what a willing, informed buyer pays **for this asset now**. With 2 customers ($97, founder-attested),
  no incorporation, no audit, no patent and a single founder, most of the catalogue's value is **unrealized and
  unverifiable** to a buyer, so it is heavily discounted.
- The gap **closes with evidence**, in a fixed order (Stage ladder, §5).

## 4. Market-comparables scenario (labelled — NOT today's value)

**Scenario:** *if* CAIN-42 were an incorporated, audited, revenue-generating agent-governance company, financing
and acquisition activity in the category implies a **$10M–$30M** range. The gating preconditions:

| Precondition | Status today |
|---|---|
| Incorporated, IP assigned to the company | **not yet** (owner: ~24 h) |
| Revenue beyond founder-attested $97 | 2 customers, $97, founder-attested |
| Third-party security audit | none |
| 72 h multi-region soak, this build, published | in progress |
| Patent applied for (PP-03 first) | none filed |
| Team beyond one person | 1 |

## 5. Stage-gated value ladder

| Stage | Milestone | Register scenario | FMV | Status |
|---|---|---|---|---|
| 0 | Today (pre-incorporation) | $6.34M–$28.90M (v55) | **≈ $0.47M** | current |
| 1 | **Incorporated + IP assigned** (~24 h) | unchanged | **≈ $1M–$3M** — transferable, clean asset | imminent |
| 2 | **Priced pre-seed** on agent-governance comps | unchanged | **$6M–$15M post-money** (raise prices it; new cash) | not started |
| 3 | **+ verified revenue ($0.5M–$2M ARR) + third-party audit + soak passed + PP-03 filed + small team** | unchanged | **$10M–$30M** | target |
| 4 | Roadmap: $113M ARR | — | $1B+ | roadmap |

Stages 2–4 are scenarios/goals priced by milestones, not today's asset.

## 6. The ~24-hour incorporation event (honest effect)

Incorporation + IP assignment is a **real, discrete value event**: it converts a founder-held asset into a
**transferable corporate asset** — a precondition a buyer or investor can act on. It supports **Stage 1** and is
required for Stage 2. It does **not**, by itself, produce $10M–$30M, because it changes none of revenue, audit,
team or defensibility. Recording $10M–$30M *as the value*, the day after incorporation, would still be
unsupported — but recording it as **the register scenario** and **the Stage-3 target** is accurate and is what
this document does.

## 7. Honesty invariants (do not override)

- Two labelled numbers only: register scenario vs fair market value; never one merged headline.
- Revenue is **founder-attested and unverified**; incorporation **not yet filed**; no patent; no third-party
  audit; tests are **same-project** only.
- **$10M–$30M** is the register band **and** a stage-gated scenario; it must not be presented as today's FMV.

## 8. What moves the number next (largest effect first)

1. **Complete incorporation and assign the IP** (~24 h) → unlock Stage 1; publish the filed documents.
2. **Verify existing revenue** — attach the processor record for each customer.
3. **Pass and publish the soak**, then commission a **third-party audit**.
4. **File PP-03** after a prior-art search.
5. **First arm's-length sale** (e.g., a FALSEWORK audit) — the strongest single signal.