An AI risk assessment template is only as useful as the fields it captures, and the fields that matter are the ones an auditor or regulator expects to see: not just the technical risk, but its impact on the people the AI system affects. A template that treats an AI risk like any other IT risk misses the element that AI governance frameworks specifically require, which is the assessment of impact on individuals and society. This guide sets out the fields an AI risk assessment template should contain, a scoring approach that prioritizes correctly, and a structure aligned to ISO 42001 and the NIST AI Risk Management Framework.
The reason to build the assessment around the right fields, rather than reach for a generic risk register, is that AI risk has a dimension ordinary risk assessments do not capture. An AI system can create risks to fairness, safety, and the rights of the people it affects, and the frameworks that govern AI expect those to be assessed explicitly. A template that prompts for them turns an AI risk assessment from a repurposed security exercise into one that actually satisfies what AI governance requires.
What an AI Risk Assessment Template Should Capture
A strong template captures each risk in enough structured detail to assess, prioritize, and act on it, without becoming so heavy that no one completes it. At minimum it should identify the AI system or use case being assessed, describe the context and purpose, state the specific risk, and capture the AI-specific dimension of how that risk could affect individuals. It then records the likelihood and severity that produce a rating, the controls already in place, the resulting risk level, and the treatment plan with an owner. Each field does a distinct job, and the assessment is only as good as its weakest field.
The field that distinguishes an AI risk assessment from a generic one is the impact on individuals and society. Ordinary risk assessments focus on impact to the organization, but AI governance frameworks require an organization to consider how an AI system affects the people subject to it, which is the substance of an AI impact assessment. Building that field into the template rather than bolting it on afterward is what makes the assessment defensible under AI governance scrutiny.
The Fields That Matter
The table sets out the fields a defensible template should contain, what each captures, and why it matters. It is a structure to adapt to the organization’s AI systems, not a fixed form.
| Field | What it captures | Why it matters |
|---|---|---|
| AI system or use case | The specific system or application being assessed | Scopes the assessment to one clear subject |
| Context and purpose | What the system does, for whom, and in what setting | Frames how the risk should be judged |
| Risk description | The specific risk, from security to fairness to safety | The core of the entry |
| Impact on individuals | How the risk could affect the people the system touches | The AI-specific element frameworks require |
| Likelihood | How likely the risk is to occur | One half of the risk score |
| Severity | How damaging the risk would be if it occurred | The other half of the risk score |
| Existing controls | What already mitigates the risk | Determines the residual risk |
| Risk rating | The resulting priority level | Drives the order of treatment |
| Treatment and owner | The plan to address the risk and who owns it | Turns assessment into accountable action |
The pattern across the fields is that the first four define and contextualize the risk, the middle three score it, and the last two convert it into action. A template missing the impact-on-individuals field fails the AI-specific requirement, one missing existing controls overstates risk by ignoring mitigation already in place, and one missing an owner produces analysis no one acts on. The fields that organizations most often omit, impact on individuals and a named owner, are precisely the ones that make the difference between an assessment that satisfies a framework and one that sits unused.
The Scoring Approach
The scoring approach in the template combines likelihood and severity into a risk rating, which is the mechanism that turns a list of risks into a priority order. Most templates use a simple matrix in which likelihood and severity are each rated on a scale, and their combination places each risk into a level such as low, medium, high, or critical. What matters is less the exact scale than that it is applied consistently, so that risks can be compared and the highest-rated ones addressed first.
The AI-specific refinement is that severity should account for impact on individuals, not only impact on the organization. A risk with modest operational cost but serious potential to harm or unfairly affect people should score as significant, because the frameworks weigh that impact heavily. Building this into the severity judgment, rather than scoring purely on business impact, is what aligns the template’s prioritization with how AI governance expects risks to be ranked. The output is a rated, ordered set of risks that tells the organization where to act first.
How the Template Aligns to ISO 42001
For organizations working to ISO 42001, an AI risk assessment is not optional but required: the standard’s planning clause requires both an AI risk assessment and an AI impact assessment, and a template that captures both satisfies that requirement directly. The risk fields support the risk assessment, the impact-on-individuals field supports the impact assessment, and the treatment and control fields feed the risk treatment and the Statement of Applicability that ISO 42001 expects. A template built around these fields therefore produces exactly the documented information an ISO 42001 audit looks for.
The connection to the wider standard is covered in the guide to ISO 42001 requirements, which shows how the risk and impact assessments sit within the management system, and the specific design of the impact assessment is covered in the guide to designing the AI impact assessment for ISO 42001. The AIMS Manual provides the broader documented-information structure the template fits into. Aligning the template to the standard from the start avoids reworking assessments later to fit.
How the Template Aligns to NIST AI RMF
The NIST AI Risk Management Framework organizes AI risk work into four functions, Govern, Map, Measure, and Manage, and an AI risk assessment template supports the Map and Measure functions in particular. Mapping is where an organization establishes the context and identifies the risks of an AI system, which the template’s system, context, risk, and impact fields capture. Measuring is where those risks are analyzed and rated, which the likelihood, severity, and rating fields support. The treatment and owner fields then feed the Manage function, where risks are prioritized and acted on, while Govern is the overarching function the whole assessment operates within.
Because the template maps cleanly onto these functions, an organization using the NIST framework can adopt it without translation, and one working to both NIST AI RMF and ISO 42001 can use a single assessment to serve both. The relationship between the two frameworks is covered in the guides to the NIST AI Risk Management Framework and to NIST AI RMF versus ISO 42001, which help teams decide how to run the two together rather than twice.
A Worked Example
A single worked entry shows how the fields fit together. Consider an AI system used to screen job applicants. The AI system field names it as the resume-screening model; the context field records that it ranks applicants for recruiter review across all open roles. The risk description captures that the model could systematically disadvantage qualified candidates from particular groups, and the impact-on-individuals field, the one that makes this an AI assessment, records that affected people could be unfairly denied employment opportunities, a serious harm the frameworks weigh heavily.
Scoring then proceeds: likelihood is judged against how the model was trained and tested, severity is rated as high because the harm to individuals is significant, and together they produce a high or critical rating even if the direct cost to the organization is modest. The existing-controls field records what is already in place, perhaps bias testing before deployment and human review of the model’s rankings, which reduces the residual risk from the inherent level. The risk rating reflects that residual position, and the treatment-and-owner field records the plan, such as expanding bias testing and adding monitoring, with a named owner accountable for it.
Walked through this way, the entry demonstrates the template’s logic: the impact-on-individuals field is what elevated a risk that a purely business-focused assessment might have rated low, the existing controls kept the assessment honest about residual rather than inherent risk, and the owner turned the analysis into action. A template that produces entries like this, specific, scored on impact to people, and owned, is one that both satisfies the frameworks and actually reduces AI risk rather than merely documenting it.
How to Use an AI Risk Assessment Template
A template becomes a real assessment through use, and the practical approach is to run it per AI system or significant use case rather than once for the organization as a whole, because risks differ sharply between systems. For each, work through the fields in order, being specific rather than generic, since a risk described as system failure gives no one anything to act on while a specific description of how and whom it could affect does. Rate consistently, record the controls honestly to reflect residual rather than inherent risk, and assign a real owner to each treatment.
The assessment should also be a living document, revisited when an AI system changes materially or a new one is introduced, because AI risk shifts as systems and their uses evolve. Treating it as a one-time exercise produces a snapshot that ages quickly, whereas revisiting it keeps the organization’s view of its AI risk current. A template gives the structure and prompts; the substance comes from applying it honestly to real systems, and from keeping it up to date as those systems change. One caveat matters: a template is a working structure that makes an assessment consistent and defensible, not a guarantee of ISO 42001 conformity or a substitute for professional advice, so the quality of the thinking entered into each field is what determines whether the output holds up.
Conclusion
An AI risk assessment template is defensible when it captures the fields that matter, the AI system and its context, the specific risk and its impact on individuals, the likelihood and severity that produce a rating, the existing controls, and the treatment and owner, with the impact-on-individuals field being the one that distinguishes an AI assessment from a generic risk register. A consistent scoring approach that weighs impact on people turns the fields into a priority order, and a structure aligned to ISO 42001 and the NIST AI Risk Management Framework means a single assessment serves both.
A template supplies the structure; the substance comes from running it per system, scoring consistently, and keeping it current as AI systems change. To build an AI risk assessment on a structure aligned to ISO 42001 and NIST AI RMF, explore Elevate’s AI governance resources or book a call with an Elevate advisor.
Key Takeaways
An AI risk assessment template is defensible only when it captures the fields AI governance frameworks require, especially impact on individuals.
- The AI-specific field is impact on individuals: ordinary risk assessments focus on impact to the organization, but AI frameworks require assessing how a system affects the people it touches, and a template must prompt for it.
- The fields split into three jobs: the first fields define and contextualize the risk, the middle ones score it by likelihood and severity, and the last ones convert it into owned action.
- Scoring should weigh impact on people: severity must account for potential harm to individuals, not only business impact, so the priority order matches how AI governance expects risks to be ranked.
- The template serves ISO 42001 and NIST AI RMF at once: it satisfies ISO 42001’s required risk and impact assessments and maps onto the NIST framework’s Map and Measure functions, so one assessment covers both.
- Adapt and maintain it: run the template per AI system, be specific, assign real owners, and revisit it when systems change, because a one-time assessment ages quickly.
FAQs
Q1. What fields should an AI risk assessment template include? A defensible AI risk assessment template should include the AI system or use case being assessed, its context and purpose, a specific risk description, the impact on individuals the risk could cause, the likelihood and severity that combine into a rating, the existing controls that mitigate it, the resulting risk rating, and a treatment plan with a named owner. The first fields define and contextualize the risk, the middle ones score it, and the last convert it into action. The field that distinguishes an AI assessment from a generic one is impact on individuals, which AI governance frameworks specifically require.
Q2. How is an AI risk scored in a template? Most AI risk assessment templates combine likelihood and severity into a risk rating using a matrix, where each is rated on a consistent scale and their combination places the risk into a level such as low, medium, high, or critical. What matters most is consistent application, so risks can be compared and the highest addressed first. The AI-specific refinement is that severity should account for impact on individuals, not only on the organization, so a risk with modest operational cost but serious potential to harm or unfairly affect people scores as significant, matching how AI governance frameworks weigh impact.
Q3. Does an AI risk assessment template meet ISO 42001 requirements? It can, if built correctly. ISO 42001’s planning clause requires both an AI risk assessment and an AI impact assessment, and a template that captures the risk fields plus the impact-on-individuals field supports both directly. Its treatment and control fields feed the risk treatment and the Statement of Applicability the standard expects. A template built around these fields produces the documented information an ISO 42001 audit looks for, which is why aligning it to the standard from the start, rather than reworking assessments later, is the efficient approach.
Q4. How does an AI risk assessment template relate to NIST AI RMF? The NIST AI Risk Management Framework organizes AI risk work into four functions: Govern, Map, Measure, and Manage. An AI risk assessment template supports Map, where context and risks are identified, through its system, context, risk, and impact fields, and Measure, where risks are analyzed and rated, through its likelihood, severity, and rating fields. The treatment and owner fields feed Manage, and the whole assessment operates within Govern. Because the template maps cleanly onto these functions, an organization can use it under the NIST framework without translation, and one working to both NIST AI RMF and ISO 42001 can use a single assessment for both.
Q5. How often should an AI risk assessment be updated? An AI risk assessment should be a living document, revisited whenever an AI system changes materially or a new system is introduced, because AI risk shifts as systems and their uses evolve. A model retrained on new data, a system applied to a new purpose, or a change in the people it affects can all change the risk picture. Treating the assessment as a one-time exercise produces a snapshot that ages quickly and no longer reflects the organization’s actual AI risk, whereas revisiting it on change and on a regular cadence keeps it current and credible.