01 / Know what you use. Know who owns it.
An AI system can arrive through a supplier, an internal project or an everyday business tool. The difficult questions soon follow: What is it used for? Who is responsible? What could go wrong? Which safeguards are in place? Where is the evidence?
Coach2Lead AICMS brings those answers together in a managed compliance dossier. A dossier is the organised record for an AI system: its purpose, responsibilities, assessments, supporting documents and review history.
It gives business owners, compliance teams and specialists a common place to work. The same record follows a system from its first assessment through approval, later changes and ongoing follow-up.
- RegisterPurpose and owner
- AssessRole and risk
- DocumentSafeguards and evidence
- ReviewIndependent decision
- MaintainChange and follow-up
A practical translation of the AI Act
Many requirements become concrete questions, linked records and checks. A missing justification can stop a dossier entering review. A reviewer can compare a proposed change with the last approved version. A risk can be linked to the people it affects, its owner and the measures intended to reduce it.
The result is a clearer account of what your organisation has considered, what it has decided and what still needs work.
AICMS helps you organise and demonstrate your compliance work. Your organisation remains responsible for the quality of its decisions, evidence and real-world safeguards.
Legal context: the EU AI Act sets different obligations according to the system, intended use, operator role and applicable dates. AICMS supports that work; it does not certify an organisation or an AI system. Regulation (EU) 2024/1689, read with Regulation (EU) 2026/1744.
02 / See the connection between a claim and its evidence
Start with the actual use
For a recruitment tool, the record can connect its intended use to the people affected, the classification assessment, discrimination risks, human oversight and the staff responsible for operating it. Guided assessments help the team work through the relevant questions.
For a high-risk AI system built around a supplier's general-purpose model, the dossier can retain model information, documentation for integrating the model and supplier evidence. Your team must still distinguish its responsibilities from those of the model provider.
These are illustrative situations, not automatic legal classifications or customer case studies.
- A decisionWhy this role and risk?
- A safeguardWhat reduces the risk?
- Its evidenceWhat supports the claim?
Review a change with its context
When a system changes, create a new draft from an existing version. Update the relevant information and compare it with the previous approved record. The review queue brings together the proposed version, change information and assessment context.
The person who created the draft or submitted it for review cannot approve it. Approval records a comment and publishes the dossier internally; the previous published version moves into the history.
Keep the follow-up attached to the system
Record periodic risk reviews, monitoring findings, incidents, corrective actions and communications with authorities. Dashboards highlight recorded attention items. Scheduled routines support reminders where the service, recipients and delivery channels are configured.
What this helps avoid: assessment documents becoming detached from the version they describe, approvals losing their rationale, and follow-up work losing its owner.
03 / A broad working record, inside Coach2Lead
The depth is in the connections. AICMS brings together the regulatory record and the people, responsibilities and evidence needed to maintain it.
| Your team's question | What AICMS brings together |
|---|---|
| What AI do we use? | A searchable register, purpose, lifecycle, classification and responsible parties. |
| Which questions should we examine? | Guided classification, fundamental-rights assessment, data governance, oversight, conformity, registration, quality management and general-purpose AI records. |
| What could harm people? | Risks, affected groups, safeguards, decisions about the risk remaining after safeguards, and review schedules. |
| What supports our claims? | Typed documents, evidence attachments, testing descriptions, training records, certificates and references to standards. |
| Who agreed to the decision? | Separate permissions, independent review, comments, version comparisons and recorded lifecycle events. |
| What needs attention now? | Portfolio views and recorded certificate, incident, review, retention and authority-response attention items. |
| What can we share? | Registry spreadsheets, printable dossiers and structured technical-documentation or registration preparation files. |
Reuse the organisation you already manage
Responsible-party and oversight records can connect to existing Coach2Lead people. Training records can reference training certificates. Safeguards can link to controls already recorded against organisational standards in Coach2Lead. This keeps AI governance connected to the organisation doing the work.
Choose the right depth
The Feature Guide explains the customer journeys. The Controls Reference explains checks and their effects. The Responsibility Map relates product support to the Act, including areas where external action or specialist judgement is essential.
These document levels are reading levels, not commercial licence tiers.
04 / Put the tool around a responsible team
A useful dossier needs knowledgeable contributors. AICMS can require an answer, preserve it and route the record through review. Your team must establish that the answer is accurate and the underlying work is adequate.
- AICMS organisesQuestions, checks, evidence and review
- Your organisation actsJudgement, technical controls, reporting and registration
What a sensible starting engagement looks like
- Choose representative AI systems. Include the intended use, suppliers, people affected and business owner.
- Agree roles and scope. Identify the organisation's legal role, the questions requiring specialist advice and the applicable rollout dates.
- Assemble evidence. Gather instructions, assessments, test results, training, agreements and existing governance procedures.
- Review a complete journey. Register, assess, document, check, submit and independently review a dossier.
- Plan continuing work. Set review ownership, incident procedures, retention arrangements and external reporting responsibilities.
Clear product boundaries
AICMS manages the governance record. It does not run safety or fairness tests on the AI, implement its human override, monitor its live outputs, apply a CE mark or submit reports to an authority. The detailed guides identify further limits, including which advisory checks are not displayed in the current editor.
Edition and availability. This bundle describes the AICMS implementation reviewed on 29 September 2026. It is a capability description, not confirmation that a particular customer environment has been deployed or configured. Production availability and enabled capabilities are confirmed during onboarding.
The conversation to have: bring one real system and ask the team to show how its purpose, risks, evidence, approval and next review stay connected.