Skip to main content

    Executive guide

    AI risk for boards: describe the event and its consequence

    A risk entry should help the board understand what could happen, whom it could affect and what management is doing about it. Broad technology categories are a starting point for that assessment.

    Connect the risk to a use

    Identify the workflow, decision, information and systems involved. A general concern about inaccurate output becomes useful when tied to a consequence, such as a customer receiving incorrect service information.

    Keep the scenario specific enough to assign ownership and test a control.

    Examine the path to harm

    Ask:

    • what would cause the event
    • whether a person would detect it
    • how far it could spread before anyone intervenes
    • whether the outcome can be corrected.

    Likelihood and severity need evidence and context. Record uncertainty, and avoid precise numbers based on weak information.

    Ask what each control does

    A policy may define permitted use. An access restriction may limit exposure. Evaluation may identify a failure pattern. Human review may prevent an incorrect action, provided it is meaningful.

    These controls address different parts of the scenario. A list of controls is not proof that all risks are managed.

    Consider dependencies

    Vendors, information sources, internal experts and integrations can become critical to operation. Establish how the business would continue if one became unavailable or unsuitable.

    A fallback needs people with the capacity and knowledge to run it. A document that says 'manual process' is not enough.

    Distinguish treatment and acceptance

    Management may remove the use, reduce the exposure, improve controls or accept the remaining risk through an approved governance process. Some requirements apply however much risk the organisation is willing to accept.

    Qualified specialists should assess applicable obligations and technical questions.

    Report the decision required

    Show the scenario, evidence, current controls, remaining uncertainty, owner and proposed action. Escalate material incidents or changes through the appropriate process as they occur. Do not wait for a routine reporting date.

    Use the AI risk register template and read the questions boards should ask about AI.

    Frequently asked questions

    1. How should a board describe an AI risk?

      Describe it as a specific event with a consequence, such as client data pasted into a public AI tool. Name the owner and the control meant to prevent it.

    2. Is a list of controls enough to manage AI risk?

      A list shows what exists, and the board also needs evidence that each control works on its scenario. Ask management for test results or incident data on the controls that matter most.

    3. When should AI risks reach the board?

      Material incidents and changes should reach the board as they occur, through the escalation route it has agreed with management. Lower-level risks can wait for the routine report.

    Complete the AI risk register template for one current AI use. Bring the rows you cannot fill in.