AI Accountability
Assessment
You are already using AI. The question is whether you know what it is doing on your behalf. One structured engagement — clear outcomes, written summary, direct follow-up. You are further along than you think.
Not a software subscription. Not a tool license. A structured engagement that replaces the hours a board, HR director, or operations manager would spend trying to audit their own AI exposure without the expertise to do it well.
Structure follows function. The person asking these questions is already further along than they think.
The question most organizations are asking is did the AI say something wrong. The question they should also be asking is did the AI do something unauthorized. That distinction is where real accountability lives — and most organizations have no framework for the second question yet. Chris does.
Action auditing is the practice of verifying not just what an AI outputs, but what actually happens in the real world because of it. It closes the gap between a model's response and the multi-step workflow or decision it triggers — tracking execution, human handoffs, outcomes, and risks along the full chain.
Action auditing evaluates not just what an AI system says, but what it is authorized to do, what actions it takes, and whether those actions remain observable, reversible, and accountable.
Every question, concern, or anxiety a client brings cannot be anticipated in advance. What can be built is the quality of the process — and the compounded evidence that the process works.
In isometric training, the nervous system limits available strength not out of weakness but out of protection. It releases more capacity only after accumulated evidence that the effort is safe. The same dynamic applies to AI governance conversations. The prospect who has been burned by an AI vendor, or who has seen a suitability complaint, or who watched a procurement decision go wrong — that person is not being unreasonable. Their cortical inhibition is appropriate. The response is not a louder pitch. It is a demonstration. The LCGE catch. The oil inversion correction. The rate lock scenario that passed output auditing and failed action auditing. Each one reduces the inhibition a little more.
The process also accommodates if it does not vary. A prospect who has heard the same AI governance framing from every advisory firm develops the same tuning-out response the nervous system develops to a repeated stimulus. Pertinent's variation is the three-verifier synthesis and the action-level audit frame — a different approach, not a louder version of the same one. Kaizen applied to governance: small, consistent improvements compounded over time.
What we work through together
- What AI tools are you currently using — and what can they do without your knowledge or approval?
- Who in your organization is accountable when an AI tool makes a mistake or produces misleading output?
- What data are your AI tools accessing, and do you know where that data goes?
- Are your current AI workflows consistent with your reporting obligations — to funders, accreditors, or governing bodies?
- Where is human judgment still essential in your workflow, and is it protected?
- What actions can your AI tools take autonomously — and do you have a record of what they did?
- A plain-language summary of your current AI exposure — what it said, what it did, and what you didn't know
- Three concrete next steps ranked by impact and effort
- A direct line back to Chris for 30 days if questions surface as you implement
Already using AI in the classroom. Needs a framework for policy, student use, and institutional accountability that holds.
Operating with AI tools and funder relationships that require clear accountability. The board question is coming.
Moving fast with AI tools. The assessment catches what the speed misses — before it becomes a problem.
Professionals whose client relationships depend on trust. Knowing what your tools are doing is no longer optional.
Every assessment is conducted personally by Chris McClean. No intake forms routed to an assistant. No automated sequences. You get the hours back you would have spent figuring this out yourself — plus a written summary you can share with your team or board.
How it works
One session. Clear accountability. A written summary you can act on.
The assessment works because you already know something is happening. This is how you find out exactly what.
Start the conversation