Free scorecard and tool · From Episode 9

AI Exit Readiness Scorecard: score your AI governance against the records a buyer’s team asks for

A buyer’s diligence team, a lender, an insurer or a new director reads your AI governance on paper. An owner who is not written down, or a tested path with no record, counts for nothing in a data room. This scorecard shows what your governance is worth when someone from outside the company comes asking.

Score 24 criteria across the six dimensions this scorecard uses to assess AI diligence readiness: AI system inventory, investment and return evidence, decision accountability, governance and control posture, operational resilience, and data and IP ownership. The tool gives you a readiness band, runs the guardrail on decision accountability and data ownership, and shows which of three conditions is holding you back.

1. A named ownerSomeone specific holds the decision, and it is written down.
2. A tested pathThe stop, the escalation and the fallback have been exercised, not just described.
3. Evidence an outsider can checkThe first two can be verified from the documents alone, without you in the room.

The owner takes the call. The path gets it to them in time. The record proves both, to someone who was not there.

Score it now

How to score: score each criterion on the evidence you can show today. 4 documented, current and evidenced over time. 3 documented and operating, minor gaps. 2 partial, informal or out of date. 1 absent, or cannot be shown. ★ marks a criterion that is material if absent.
Evidence levels
Finance validated: finance has confirmed the baseline and the calculation.
Operationally verified: measured by the operating team against a recorded baseline.
Directional: estimated or observed, not measured against a baseline.
An independent audit or attestation is stated separately where one exists.

1. AI system inventory and documentation

A complete, current inventory of every AI system in operation, with named owners, risk classifications and governance status.

★1.1The inventory covers every AI system in operation, including AI embedded in ERP, CRM and other SaaS platforms.Evidence: The inventory itself
1.2The inventory has been updated within the last 90 days.Evidence: Last review date on the inventory
1.3Each system has a named owner who has signed off the entry.Evidence: Owner sign-off for each entry
1.4Each system carries a risk classification reviewed within the last 12 months.Evidence: Classification record with review date
Dimension 1 score: not yet scored

2. Investment and return evidence

Documented evidence that AI investments returned measurable value, in terms that would survive CFO or auditor scrutiny in a transaction.

2.1Each material AI initiative has a documented business case.Evidence: Approved business case for each initiative
2.2Outcome metrics are tracked and reported, not activity metrics such as uptime or adoption.Evidence: Outcome reporting for each system
2.3Return evidence reaches operationally verified for most material systems.Evidence: Evidence rating for each system
★2.4Any AI value claimed in deal materials is finance validated. Any other figure appears only with its evidence level stated: operationally verified or directional.Evidence: Value bridge with finance sign off on the baseline and the calculation, and an evidence level on every figure
Dimension 2 score: not yet scored

3. Decision accountability and reconstructabilityGuardrail

For material AI-influenced decisions, the company can show who approved what, using which output, with what override rights, and what the outcome was.

★3.1Decision rights are documented for every material AI-influenced decision, in a register updated within the last 6 months.Evidence: Decision rights register with review date
3.2An override protocol is in place and has been tested for each high-risk system.Evidence: Override protocol and test or override log
★3.3One material decision can be reconstructed end to end: approver, model output, override rights and outcome.Evidence: One reconstructed decision record
3.4An evidence retention policy is active and can be audited.Evidence: Retention policy and retention periods applied to logs
Dimension 3 score: not yet scored

4. Governance and control posture

Core governance controls are in place, documented and operating: AI use policy, incident response, board reporting and regulatory exposure assessment.

4.1A current AI use policy has been approved by the board.Evidence: Policy with approval and review dates
4.2An incident response protocol exists and has been exercised.Evidence: Protocol and a tabletop or incident record
4.3AI appears in board reporting with a named executive owner.Evidence: Most recent board presentation on AI
★4.4Regulatory exposure has been assessed for every AI regime that applies to the company.Evidence: Exposure assessment reviewed with counsel
Dimension 4 score: not yet scored

5. Operational resilience

Critical AI systems have fallback procedures, vendor dependencies are documented, and operations can continue if a key system becomes unavailable during or after the transition.

5.1A fallback procedure is documented and tested for each critical AI system.Evidence: Fallback procedure and last test date
5.2A vendor dependency summary covers third-party AI and model dependencies.Evidence: Vendor dependency summary
5.3AI vendor contracts have been reviewed for change of control and termination provisions.Evidence: Contract review notes
5.4Knowledge of how each critical system runs is held by more than one person.Evidence: Runbooks and a named second person
Dimension 5 score: not yet scored

6. Data and IP ownershipGuardrail

Data used in AI systems is owned, licensed or contracted in a way that transfers cleanly in a transaction, and third-party model dependencies are understood.

6.1Data provenance is documented for each AI system.Evidence: Provenance record for each system
6.2Third-party model dependencies are listed with their license terms.Evidence: Model dependency list with license terms
★6.3Third-party data and model licenses are confirmed to transfer on a change of control.Evidence: License review confirming transferability
6.4No proprietary third-party data is used without disclosure.Evidence: Data use attestation from each system owner
Dimension 6 score: not yet scored

Your result

Score all 24 criteria to see your readiness band, guardrail check and the condition holding you back.
Get the full scorecard

Scores stay in your browser. Nothing is saved or sent. A self-score is not independent validation; a buyer’s team will test the evidence behind it.

Get the full scorecard

The full version adds what a strategic buyer, a financial sponsor and counsel will each ask on every dimension, an exit timeline for what to build, evidence or disclose at 24, 12 and 6 months, a 20-item data room list, and a one-page sponsor summary for the operating partner or board. PDF and Word. The links arrive by email.

Name

Want it scored by someone a buyer will believe?

The AI Diligence Readiness Review applies the full VTCDO review to your company: interviews, an evidence review across all six dimensions, a reconstructed decision on each material system, scoring with written rationale, a remediation tracker and a data room list. $30,000 fixed, delivered in 15 business days, for up to five material AI systems, or $22,500 for up to two. Pre-exit or buy-side.

How the Diligence Readiness Review works →   Start a conversation →

A self-assessment and operating tool, not legal advice. Whether a legal duty applies, and what must be disclosed in a transaction, belongs to your counsel. Built on the VTCDO AI Diligence Readiness Review.