AI Tool Usage Tracking: Why CASB & DLP Miss Shadow AI

A consultant-grade ISO 42001 gap analysis checklist has 38 Annex A controls, roughly 80 clause-level “shall” statements, and one question attached to every line: where is the evidence, and would a certification body accept it? That last question is what separates the checklists consultants use from the free self-assessment spreadsheets that rank for the same search. This article lays out the checklist itself: what a consultant checks before the engagement starts, the clause-by-clause and control-by-control checkpoints, how evidence gets sampled, how gaps get scored, what the deliverables look like, and what fails most often. Use it to run your own assessment, or to check whether the consultant you’re about to hire is doing the job properly. What Makes a Consultant-Grade ISO 42001 Gap Analysis Checklist Different​ Depth of Evidence Review vs. Self-Assessment Tools A self-assessment tool asks whether you have an AI policy. A consultant asks to see it, checks the approval date and version, reads clause 5.2 against it, and then asks three people in engineering whether they’ve read it. The checklist item is the same. The evidence standard is not. Consultants score every item on three levels: documented, implemented, and effective. A policy that exists but nobody follows scores as “ad hoc,” not “defined.” A control that runs but produces no record scores as unverifiable, which for audit purposes is the same as absent. Self-assessment tools collapse those three levels into a single yes/no, which is why companies that score 85% on a free tool routinely receive major nonconformities at Stage 2. Alignment with Certification Body Expectations Certification bodies auditing against ISO/IEC 42001:2023 now work under ISO/IEC 42006:2025, which sets competence, audit-time, and impartiality requirements for AIMS auditors and builds on ISO/IEC 17021-1. A consultant-grade checklist is written with 42006 in mind: it organizes findings by clause and control identifier, because that’s how the auditor works, and it records evidence locations, because that’s what the auditor will sample. The practical difference shows up in the report. A gap register that says “AI governance needs improvement” is useless in front of an auditor. One that says “A.5.2 not conformant: no documented impact assessment process; two of four in-scope systems have no assessment on file” maps directly to the audit plan. Risk-Weighted Scoring Methodology Self-assessments count gaps. Consultants weight them. A missing AI policy under clause 5.2 and an incomplete competence matrix under 7.2 are both gaps, but the first will block certification and the second will earn you a minor finding. A consultant-grade checklist carries two scores per line: a maturity rating (how far the control is from working) and a certification criticality (what happens at audit if it stays this way). Effort estimates live in the remediation plan, never in the gap score, because mixing them produces a roadmap that fixes easy things first rather than important ones. Insider Note: The fastest tell that a checklist is consultant-grade rather than a marketing download is whether it has a column for evidence location. Auditors don’t accept “yes” as evidence. If the checklist has nowhere to record where the proof lives, it wasn’t built by someone who has sat through a Stage 2. Pre-Engagement Preparation Consultants Complete Before the Gap Analysis Client AI Inventory and Use Case Cataloging Nothing in the checklist works without a complete AI inventory, and it’s the input clients get wrong most often. The inventory records every AI system in use: purpose, the role you play (developer, provider, deployer, or user), data consumed, outputs produced, whether a human sits between the output and the decision, and which third-party model or API it depends on. Consultants push hard on shadow AI here: SaaS tools that added AI features, agents running under employee credentials, and internal scripts calling model APIs. Every one of those is in scope until you document why it isn’t. Defining AIMS Scope Boundaries Clause 4.3 requires a scope statement naming which AI systems, business units, locations, and lifecycle stages the AIMS covers. Consultants draft this from the inventory, not before it. Scope discipline matters commercially too: certification bodies price audits by audit days, and audit days scale with scope. A narrow, well-justified first scope (the customer-facing AI product, say, rather than every internal tool) is usually the right call for a first certification. Stakeholder Interview Planning The checklist needs answers from people who don’t write policies. A typical interview plan covers the executive sponsor (clause 5), the AI or product lead (clauses 6 and 8), data engineering (A.7), procurement or vendor management (A.10), legal or privacy (A.5, A.8), and at least one front-line user of the AI system (A.9). Consultants interview the doers separately from the document owners, because the distance from what the procedure says to what actually happens is the finding. Document Request List (DRL) Consultants Send Clients The DRL goes out one to two weeks before fieldwork. A standard ISO 42001 DRL asks for the AI inventory; existing AI, security, and data policies; org chart with AI governance roles; any AI risk assessments or impact assessments; model documentation (model cards, system cards, or whatever exists); training-data provenance and data quality records; supplier contracts for third-party models; incident and change logs; training records; any ISO 27001 ISMS documentation; and the last internal audit and management review minutes if they exist. Missing items become findings rather than delays. Pro Tip: Return an Honest DRL Return the DRL with a column that says “does not exist” wherever that’s true. Consultants would rather know on day one than discover it in a workshop. An honest DRL shortens fieldwork by days and makes the maturity scores more accurate, which makes the remediation plan cheaper. Clause-by-Clause Checklist Consultants Use (ISO 42001 Clauses 4 to 10) ISO 42001 follows the Harmonized Structure shared with ISO 27001 and ISO 9001, so clauses 4 to 10 will look familiar to anyone who has run an ISMS. What’s different is the content each clause demands. Clause 4 – Context of the Organization Checkpoints Consultants check for a documented analysis of

ISO/IEC 42001:2023 asks for three assessments, and most teams try to squeeze them into one spreadsheet: a gap analysis against clauses 4 to 10 and Annex A, an AI risk assessment under clause 6.1.2, and an AI system impact assessment under clause 6.1.4. Treat them as one exercise and the auditor pulls them apart for you at Stage 2. Treat them as three unrelated projects and you triple the workshops, the registers, and the remediation lists. What works is a single methodology with distinct outputs that share inputs, share a traceability matrix, and feed one remediation plan. This article lays out that methodology end to end: how gap analysis and risk assessment fit together under ISO 42001, how to prepare, the step-by-step process for each, how to merge the outputs into one risk treatment plan, the registers and templates you’ll need, and what a certification body expects to see when you’re done. Why Gap Analysis and Risk Assessment Must Work Together Under ISO 42001 A gap analysis measures distance from the standard. A risk assessment measures exposure from your AI systems. They answer different questions, and ISO 42001 makes them depend on each other in a way ISO 27001 only implies. Clause 6.1.3 requires you to compare the controls you select through risk treatment against Annex A, and to justify any Annex A control you leave out in the Statement of Applicability (SoA). So your Annex A gap analysis has no defensible baseline until the risk assessment tells you which controls you need. Run the gap analysis on its own, and you end up scoring yourself against all 38 controls, including ones your risk profile never called for. Run the risk assessment on its own, and you pick treatments with no idea what already exists to deliver them. The methodology below interleaves the two. A clause-level gap review sets the scope and evidence base, the risk and impact assessments decide which controls are required, and a control-level gap review then scores only what matters. How AI-specific risks shape the methodology Traditional information security risk works from confidentiality, integrity, and availability. AI risk adds categories that don’t map neatly onto any of those: model drift, bias in training data, outputs nobody can explain, automation bias in the humans doing the reviewing, and dependence on third-party foundation models whose behavior changes without warning. ISO/IEC 23894, the companion guidance on AI risk management, adapts the ISO 31000 cycle (establish context, identify, analyze, evaluate, treat) to these sources rather than inventing a new one. That’s why the methodology here keeps the familiar ISO 31000 shape and changes the inputs, not the process. Regulatory and business drivers for a formal methodology The commercial driver is procurement. Enterprise security questionnaires now ask whether you ran an AI impact assessment, whether a human reviews high-stakes outputs, and which third-party models touch customer data. A documented methodology answers those questions with evidence instead of assurances. The regulatory driver is the EU AI Act, and its timeline moved in July. Regulation (EU) 2026/1744, the Digital Omnibus on AI, entered into force on July 27, 2026, and pushed the high-risk obligations for standalone Annex III systems from August 2, 2026 to December 2, 2027. Annex I embedded systems moved to August 2, 2028. The Article 50 transparency obligations still kicked in on August 2, 2026, as originally planned. Article 9 of the AI Act text on EUR-Lex requires a risk management system for high-risk AI that runs continuously across the system lifecycle, which is exactly what an ISO 42001 methodology gives you. Sixteen extra months is time to build it properly, not a reason to shelve it. Core Principles of an ISO 42001 Gap Analysis and Risk Assessment Methodology Four principles keep the methodology defensible in front of a certification body. Alignment with clauses 4 to 10 and Annex A. Every finding in the gap register cites a clause or an Annex A control identifier. Auditors work clause by clause, so a gap register organized any other way forces a translation step during the audit that nobody enjoys. Integration with the AI system impact assessment. Clause 6.1.4 is what separates ISO 42001 from every other Annex SL standard. The impact assessment looks outward at individuals, groups, and society. The risk assessment under 6.1.2 looks inward at the organization. The standard wants both as separate documented outputs, and the consequences you find in the impact assessment have to feed back into the risk assessment. So the methodology runs the impact assessment as a scheduled input to risk analysis, not something bolted on the week before the audit. Risk-based thinking applied to the AIMS itself. Clause 6.1.1 also asks you to consider risks and opportunities to the management system: someone leaving the AI governance function, a vendor retiring a model, a regulator changing its classification rules. These go in the same register with a different category tag. Defined inputs, outputs, and success criteria. Inputs are the AI system inventory, the scope statement, existing policies, data flow diagrams, model documentation, and your risk criteria. Outputs are the gap register, the AI risk register, impact assessment reports, the SoA, and the risk treatment plan. Success means each output traces to the others, every gap and risk has an owner, and an internal auditor could repeat the process and land somewhere similar. Insider Note: Impact assessments are where certification auditors probe hardest, because they’re the most distinctive part of ISO 42001 compared with ISO 27001. A recycled security risk register with “AI” pasted into the risk titles gets picked apart in Stage 2. Build the impact assessment methodology properly the first time. It’s far cheaper than rebuilding it under a nonconformity deadline. Preparing for the Gap Analysis and Risk Assessment Preparation is where most of the calendar time goes, and where most later problems start. Define scope, boundaries, and the AI system inventory. Scope under clause 4.3 has to name which AI systems, business units, and lifecycle stages the AIMS covers. You can’t write