Document an AI decision in under 30 minutes.
Know if your AI use is safe, risky, or unacceptable — before someone asks you to justify it.
Designed for decisions that may later be reviewed, questioned, or challenged.
When a board, auditor, regulator, or court asks how this AI decision was made — will you be able to answer?
Defensibility cannot be reconstructed after the fact. It must exist in writing, with contemporaneous reasoning, before the consequences of the decision are known. Otherwise, however sound the decision was, it will not hold under scrutiny.
Including — AI Acceptable Use Policy, Use Case Triage Tool, Shadow AI Disclosure Protocol, Vendor Review Checklist, Executed Example: Marketing Content, and a 30-day Deployment Roadmap.
Everyone whose signature — formal or de facto — stands behind a decision that might be reviewed later.
When scrutiny arrives unannounced.
A new board asks how a decision was authorized. An auditor requests the reasoning behind a deployed AI system. A regulator opens a review. A claimant's lawyer files a request. You need to show — not tell — how the decision was made. If the record does not already exist, it is too late to build one.
When the outcome goes wrong.
Some decisions, however well reasoned, produce adverse outcomes. The question you will be asked is not whether the result was perfect. The question is whether you can demonstrate that the process was diligent, informed, and proportionate at the time the decision was made. Without contemporaneous documentation, the answer collapses.
When someone else takes over your chair.
New ownership. A successor. An incoming executive team. They inherit your decisions — and the responsibility for defending them. If your records cannot stand on their own written merits, the weight of your judgment becomes their problem. Structured records protect both your legacy and their ability to move forward.
Defensibility decays over time.
Memory fades. Participants move on. Alternative options considered but rejected become unprovable. The reasoning that made sense at the moment of decision becomes indistinguishable from post-hoc rationalization. This decay is structural. It cannot be reversed by intention or diligence alone.
What remains defensible in twenty-four months is only what was written down before the outcome was known — not what was remembered, argued, or reconstructed after.
Deploying governance for your own clients? See the Agency License →
The Decision Protocol is currently in its first published edition. Organizations adopting it now form the early operator cohort — the practitioners whose application of the method will inform the next iteration.
What this is
The Triage Protocol
Every AI deployment is classified into one of three levels of documentation depth: Express (15 minutes, low-risk reversible decisions), Standard (30–45 minutes, material impact), or Shielded (60+ minutes, critical liability or regulatory exposure). Every decision takes the file it deserves — no more, no less.
The Decision Files
Each record captures what was known, what was assumed, which alternatives were considered, which trade-offs were consciously accepted, and who signed what. Every file is designed to stand on its written merits before an independent reviewer who was not present at the moment of decision.
Module A & Module B
The Bias Audit (Module A) and the Pre-Mortem (Module B) stress-test high-stakes decisions before ratification, documenting that cognitive risks and failure modes were actively reviewed — not ignored. The presence of these challenge records is itself evidence of advanced diligence.
Complete AI Governance Suite
38 structured instruments across both frameworks plus four exclusive implementation components — everything you need to deploy a defensible decision system in 30 days.
One-time purchase · Immediate digital access
Or purchase individually
AI Use & Control Framework
One-time purchase · Immediate digital access
17 files · For internal rules, triage, literacy, disclosure, and rollout control.
AI Vendor & Data Control Framework
One-time purchase · Immediate digital access
17 files · For external AI tool review, restricted-data handling, vendor approval, and escalation discipline.
Other ways to start
Starter Tool — $29
A single defensible decision record. For documenting one specific AI decision — before scaling up to a full framework.
35+ years in European institutional banking, fiduciary governance, and regulatory risk — including senior roles at ABN AMRO Bank (Amsterdam HQ), Banco Santander, Monte dei Paschi di Siena, and Mediolanum Bank. Currently Board Member of the Vatican Foundation “Sacra Famiglia di Nazareth”.
Read full bio → · LinkedIn · Foundation BoardThe method as a commercial instrument for consulting and advisory practice.
Consulting and advisory professionals who adopt the Decision Protocol gain access to a complete governance methodology they can apply within their own professional engagements — with firm-branding permissions, broad-scope deployment, and operator-grade guidance.
Governance engagements delivered by licensed advisors are typically invoiced between USD 15,000 and 35,000 per client, depending on engagement depth. The Agency License is structured for application within such engagements — providing the operational method the advisor uses across client work, reducing per-engagement preparation time, and establishing a recognizable standard of work across the advisor’s client portfolio.
Five governance actions. One coherent internal control layer.
Define internal rules
Establish what is permitted, restricted, and prohibited.
Classify use cases
Route each use case through structured triage.
Control tools and data
Evaluate vendors. Classify data. Approve or restrict.
Assign ownership
Define who approves, who escalates, who owns.
Preserve the record
Maintain review discipline and governance trail.
Six layers. One coherent structured method.
Each component has a defined role. Together they produce a structured internal control capability for defensible decisions.
Acceptable use and internal rules
Define what is permitted, what is restricted, and what is prohibited for AI use within the organization.
Vendor and data controls
Evaluate external AI tools. Classify data. Approve vendors. Control what data flows where.
Use-case triage
Route every AI use case through structured screening before it proceeds.
Ownership and escalation
Assign clear ownership. Define escalation triggers. Establish approval authority.
Documentation and examples
Maintain decision records. Provide worked examples. Support adoption with editable instruments.
Additional editions planned
Additional editions of the method may be released separately — Banking Risk, Board Decisions, and others — as part of the broader library.
The Preview is real material — not a brochure.
Actual pages, real structure, and real operating logic from the current frameworks. Enough to judge the quality before buying.
The Interactive Preview is a guided tour with a working triage tool and editable policy demo. If the Preview does not meet your standard, do not buy.
Real structured instruments. Real structure.
Not marketing copy. Not a summary. Actual pages from the instruments, with enough depth to evaluate the writing standard, underlying method, and operational usability.
Why trust this
Inspectable before purchase
Real pages and real structure are visible before payment.
Immediate delivery
Files are delivered immediately after purchase.
Clear boundaries
No consulting, legal advice, or post-purchase advisory support is included.
Operator with institutional pedigree
Founded by Lamberto Iezzi — 35+ years in institutional banking, fiduciary governance, and regulatory risk. Read the full bio →
Designed as a system
The available frameworks are standalone and usable now, but also designed as part of a broader structured method for defensible decisions.
Free sample
See what a completed decision record looks like
This is not a template. It is a completed decision record — classification, triage routing, data handling, review layers, trade-off documentation, and closure — exactly as it would appear inside your organization after deploying the framework.
7 pages · 9 sections · Marketing content use case
Open the Executed ExampleInside the record
§1 Classification & Triage
§2 Decision Context
§3 Data Sensitivity
§4 Review & Approval
§5 Decision Rationale
§6 Accepted Trade-off
§7 Conditions & Restrictions
§8 Governance & Closure
Frequently Asked Questions
These are downloadable structured instruments for defensible decision records, applied to AI deployment. Designed for immediate organizational use.
No. The frameworks are delivered as downloadable materials only. No consulting, implementation services, legal advice, certification, or post-purchase advisory support are included.
No. They are structured systems of instruments that combine policy, triage, disclosure, review logic, guidance materials, and executed examples.
They are designed for organizations already using AI that need clearer internal rules, control logic, data-handling boundaries, and the discipline to produce decision records that hold under later scrutiny.
If you need internal rules, literacy, disclosure, and triage for everyday AI use, start with the AI Use & Control Framework. If you need vendor review, restricted-data handling, onboarding, and approval logic for external AI tools, start with the AI Vendor & Data Control Framework.
Yes. The AI Edition is the first application of the Decision Protocol method. Additional editions may be released separately for other domains. Only the frameworks shown on the site are currently available for purchase.
Small teams can deploy in a single session. Mid-sized organizations typically deploy in 2–4 weeks. The Complete Suite includes a 30-day deployment roadmap and an Enterprise Deployment Checklist for organizations with 200+ employees.
These frameworks provide the practical structured instruments — policies, triage, controls, and contemporaneous records — that support organizational readiness regardless of which regulatory standard applies. The method is regulation-aware but not regulation-dependent.
Immediate digital access. You receive an email with a download link for a structured zip file containing all documents (PDF), editable templates (HTML), and interactive decision tools (HTML). Everything works offline in any modern browser.
The frameworks are designed to be adopted and customized internally without legal assistance. For regulated sectors or organizations with specific compliance requirements, independent legal review of the adapted policy is recommended before formal adoption.
The frameworks are sector-agnostic. They apply to any organization using AI tools — financial services, healthcare, legal, technology, consulting, education, or any other sector where decision defensibility is needed.
Yes. A free Preview is available so you can inspect the quality before purchase. Full refund policy details are available on the Refunds page.
Defensibility is not a future project. It is a current requirement.
Deploy a complete set of structured instruments in 30 days. Both frameworks, 20 case records, executive briefing, and a step-by-step roadmap.
Regulatory Crosswalk — how the method maps to recognized standards
For compliance, legal, risk, and advisory leaders evaluating whether the Decision Protocol method is recognized by the frameworks they answer to: a structured mapping to NIST AI RMF, ISO/IEC 42001, the EU AI Act, and SR 11-7 Model Risk Management. Free download. Two volumes plus executive summary. Honest about coverage and gaps.