Today is a build day, so this issue is shorter on theory and longer on things you can actually use. Three releases. The first is free and time limited. The second is free and permanent. The third is the book that closes the loop on the first two.
1. The AI Audit opens in beta, and the first 100 reports are free
The AI Audit is now open in beta, and the first 100 people who sign up get the full audit report free, with no restrictions. The only condition is that you tell me where it breaks.
Here is why this exists.
Most leadership teams cannot answer a simple question about their own operation: what AI is running in this business, who owns it, what it costs, what it is worth, and what happens if someone asks us to defend it. Not because the people are careless. Because AI did not arrive as a procurement decision that anyone tracked. It arrived as a feature switched on inside software the company already owned. The meeting recorder started summarizing. The CRM started drafting. The design tool started generating. Nobody approved any of it as AI, so nobody inventoried it, and now the picture is genuinely unknown rather than merely undocumented.
The audit is the instrument that closes that gap. It is not a survey, and it is not a maturity model dressed up as a quiz. It is an operational diagnostic, the same instrument used inside six-figure consulting engagements, made available directly to a single operator.
It measures five dimensions at once. Tool inventory, what is actually running, including the tools nobody registered. Cost mapping, what each one costs and who owns that cost. Performance and value, what each one delivers against what it consumes. Risk and exposure, where the data goes, what it touches, and what that creates. Governance and accountability, who is answerable, and whether that is written down anywhere.
The five run together on purpose. They are tangled together in practice, so measuring one in isolation produces a partial picture that misses the dimensions that actually matter. A tool that is not on anyone's inventory has no cost owner, no performance baseline, no risk classification, and no accountable executive. Audit only the tools, and you learn nothing about the exposure. Audit only the risk, and you cannot tell which risks attach to systems that are also producing real value.
What the first 100 get is the full audit report. Not a teaser, not top-line scores with the useful part behind a paywall. The complete report, no restrictions, the same one that costs $399 once the beta closes.
What I am asking in return is that you tell me what breaks.
That is the whole deal, and I want to be direct about it because it is the actual trade. This is a beta. Some questions will read ambiguously. Some industry classifications will not fit your business cleanly. A score may land in a way that does not match what you know to be true about your own operation. When that happens, I want to hear about it, specifically and bluntly, rather than have you close the tab and think less of the instrument.
You are getting a $399 report for nothing. I am getting something worth more than $399, which is 100 senior operators stress-testing a diagnostic against 100 real businesses before it goes out at scale. That is a better trade for me than it looks, and it is the reason the offer is what it is.
The questionnaire runs about 35 minutes, 127 questions across the five dimensions. Answer it yourself, or have someone who genuinely knows the operation answer it. This is built for a senior operator, not for forwarding to IT with a note that says "can you fill this out."
2. A new Resources page, and everything on it is free
The reference material had grown past the point where it could live scattered across the navigation, so it now has one home. Everything on it is free, maintained against primary sources, and written for people who have to defend a decision rather than describe a technology.
Three pieces are worth calling out individually.
The AI Glossary, 416 terms. https://srjconsultingservices.com/resources/ai-glossary/
Four hundred and sixteen terms across ten categories, defined in plain English, with a live search box and an A to Z filter.
This exists because of a specific failure mode I watch happen in real rooms. A vendor uses a word. Nobody in the room wants to be the person who asks what it means, so the word passes unchallenged, and a decision gets made on top of a term that half the table understood differently. "Agentic." "Fine-tuned on your data." "Guardrails." "Grounded." Each of those means something specific, and each is used loosely enough in sales conversations that the loose version has started to sound like the real one.
The categories run from the technical to the cultural. Safety and Governance is the largest at 66 terms, then Systems and Infrastructure, Training and Mechanics, Prompting and Interaction, Architecture, Core Concepts, Slang and Culture, Business and Adoption, Benchmarks and Evaluation, and Agents. Yes, Slang and Culture is a real category with 34 entries. If your team is using a word you have not heard, it is probably in there.
The AI Governance Reference Library, 61 pages. https://srjconsultingservices.com/ai-governance/
Sixty-one pages covering the frameworks, laws, agency enforcement postures, and sector rules that govern AI in the United States and the European Union, plus dedicated coverage of China and ten further jurisdictions.
The thing that makes it worth your time is not coverage. Coverage is easy to claim. It is the corrections.
The library carries verified corrections on points where the field is confidently wrong, and it says so on the page rather than quietly publishing the right answer and hoping nobody noticed the difference. Where a law has been repealed, the page says it was repealed. Where two separate regulatory instruments are being routinely conflated in compliance write-ups, the page separates them and names which one actually applies. Where a standard permits a self-assessment route that most summaries describe as mandatory third-party assessment, the page says so.
That is the editorial posture: correction over currency. Being first to publish is worth very little if the thing published is wrong, and in this field a surprising amount of what circulates is a confident restatement of someone else's error.
The AI Tools Catalog, 317 tools. https://srjconsultingservices.com/ai-governance/ai-tools/
Three hundred and seventeen AI tools across 23 categories. Each one listed with its vendor, the vendor's headquarters jurisdiction, and the governance flag it raises, from recording consent to source-code exposure to cross-border data review.
Headquarters jurisdiction is on there because jurisdiction is a governance fact, not trivia. It determines which data transfer, privacy, and cross-border rules apply to information that passes through the tool. Tools whose home jurisdiction warrants a data-flow review before business data touches them are flagged as such.
This is not an endorsement list. Presence means a tool is common enough in the field that it belongs on an inventory checklist, nothing more. Think of it as the working list to check your own inventory against, which is exactly how it gets used in practice.
Also on the Resources page: the complete bibliography behind the library, 74 primary sources and 45 peer-reviewed or formally published research works, maintained the way a book maintains its reference section. Plus Insights, this newsletter, and the press kit.
3. Volume III is available now: The AI Risk & Governance Review™
How Executives Defend Their AI Decisions When the Board, the Regulator, the Acquirer, or the Lawyer Asks.
Volume III of The Operating Discipline for AI Library™ is out.
Volume I asked what AI could do for the business. Volume II asked whether the organization was ready to absorb it. Volume III asks the question that has quietly replaced both of them in the rooms where decisions actually get made: is this program defensible if someone tests it?
Four stakeholders have arrived at the AI conversation, and they are not asking about adoption. A board member asks who signed off on the customer-facing model. A regulator asks for documentation of how a decision affecting a person was made. An acquirer's diligence team asks for an AI attestation as part of the data room. A cyber, E&O, or D&O carrier puts AI questions on the renewal application.
The question lands the same way every time, and it is not what tools are you using or how many employees have logged in. It is: what do you hand them?
Most leadership teams do not have a clean answer right now. A slide deck about AI strategy does not close it. A vendor policy stapled to the employee handbook does not close it. The gap between running AI and being able to defend running AI is what this book is built to close.
Inside it: the 6-Step Review, which produces a per-use-case governance dossier for each material AI use case. The AI Governance Framework Crosswalk™, mapping that dossier to the frameworks a regulator, auditor, or acquirer will actually reach for. The AI Accountability Matrix™, placing executive accountability so a board can see who owns what. The AI Governance Maturity Scale™, a five-dimension score that complements the readiness scoring from Volume II. The AI Data Exposure Model™, the Decision Influence Matrix™, the AI Vendor Risk Inventory™, and the AI Steering Committee Charter™. Plus three new operating logs that satisfy ISO/IEC 42001 Clauses 9 and 10.
It is written for the leader who has already done the AI work. You told the team it would change the business. You absorbed the friction of adoption. And now a different kind of question has arrived, from a different kind of person, and the honest answer is that the evidence is not assembled. The book assembles it.
How the three fit together
They are one sequence, which is why they shipped on the same day.
The audit tells you where you actually stand, in about 35 minutes, across the five dimensions that matter. The Resources page gives you the reference material to interpret what it found, the vocabulary, the frameworks, and the tool-by-tool governance flags. The book gives you the operating discipline to convert all of it into something you can hand across a table when someone asks you to defend it.
You can start anywhere. But if you have not run the audit, start there, because the full report is free right now for the first 100 and it takes about 35 minutes, and it is genuinely difficult to govern something before you have looked at it squarely.
Start your AI Audit: https://aiauditforcompanies.com/startaiaudit/
Stephen R. Jordan
Founder and Principal Advisor, SRJ Consulting & Services LLC
Author, The Operating Discipline for AI Library™
If you run the beta, reply to this email with anything that felt wrong, unclear, or off. That is the price of admission, and I read every one.