`brando:ImpactAssessment`

AI brand impact assessment record aligned with enterprise AI governance and emerging AI impact assessment practices (e.g. ISO/IEC 42005). It captures purpose, scope, brand and customer risks, mitigations, expected and observed value, and decisions for an AI system, use case, or model from a brand, communications, and CX perspective.

rdfs:comment
An assessment of the brand, customer, and regulatory impact of an AI system, use case, or model, including risks, mitigations, governance decisions, and value outcomes. Designed to complement enterprise AI, privacy, and model risk frameworks, and align with impact assessment practices such as ISO/IEC 42005.

Term definition

IRI: brando:ImpactAssessment
Type: owl:Class
Superclass: schema:Action, schema:CreativeWork
Defined in: brando: (Brando Schema Vocabulary v1.3)

brando:ImpactAssessment is the core brand-focused impact, risk, and value node in Brando. It represents a first-class assessment object that:

links to the brand, context, and AI artefacts (models, prompts, automations) being assessed,
describes intended use, brand and customer impacts, risks, value hypotheses, and mitigations,
records governance outcomes (e.g. approved, rejected, pilot-only),
can reference or sit alongside enterprise AI / privacy / model risk assessments maintained elsewhere,
provides a measurement spine for tracking success metrics and value over time for anything integrated with the Brand Oracle.

It is designed to align structurally with AI impact assessment flows such as ISO/IEC 42005, while remaining light enough to be used operationally inside a Brand Oracle, with a focus on how AI touches the brand and what value it creates rather than the entirety of enterprise AI governance.

Intended usage

Use brando:ImpactAssessment whenever you need to formally record how an AI use case or configuration has been evaluated from a brand, customer, communications, and regulatory-signal perspective, including how success will be measured.

Common patterns:

Pre-deployment brand impact & value assessment
- New assistant, tool, or feature using generative AI in a branded surface.
- New prompt flows or agents that can speak or act on behalf of the brand.
Material change assessment
- Model upgrade, new data source, new high-risk channel (e.g. wealth, healthcare, youth, public social).
Periodic / portfolio review
- Quarterly re-assessment of high-impact, high-visibility use cases.
- The Brand Brain’s “AI inventory” view: each live use case has an attached brand impact & value assessment.

This class is not intended to replace:

enterprise AI ethics / responsible AI processes,
privacy / data protection impact assessments,
model risk governance or security assessments.

Instead, it links to and references those where relevant (e.g. via brando:linkedPolicy, brando:externalSystemId, or tags), while providing a brand-governance front door:

Is this use case on-brand?
What could it do to trust, tone, and customer outcomes?
What value do we expect it to create?
How will we measure that value as it evolves?
Under what conditions is it allowed to run?

In a typical flow:

A team defines a brando:Brand, brando:Context, and relevant runtime artefacts:
- brando:BrandModel, brando:Prompt, brando:AutomationRule.
They create a brando:ImpactAssessment instance linked to those nodes.
The assessment captures:
- purpose, user groups, channels, jurisdictions;
- potential brand & customer harms and regulatory exposure as it shows up in comms;
- success criteria, metrics, and expected outcomes;
- mitigations (policies, controls, guard rails) and residual risk.
A governance body (e.g. Brand, Legal, Risk, Compliance) records a decision (approve, reject, pilot-only) and a rationale.
Runtimes and dashboards query brando:ImpactAssessment to determine:
- whether a given use case is approved,
- under what conditions,
- how it is performing against its success metrics, and
- when it must be re-reviewed.

Relationships

Inbound

Impact assessments are usually reached from:

brando:hasImpactAssessment
- Domain: brando:Brand, brando:Context, brando:BrandModel, brando:AutomationRule
- Range: brando:ImpactAssessment
- Comment (summary): Links a brand, context, model, or automation rule to one or more recorded AI impact assessments.

Example:

{
  "@id": "https://example.com/brand/northstar/models/wealth-assistant-v1",
  "@type": "brando:BrandModel",
  "schema:name": "Northstar Wealth Assistant v1",
  "brando:hasImpactAssessment": {
    "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial"
  }
}

Impact assessments will often also be linked via governance structures, for example:

brando:linkedPolicy → brando:Policy
future governance types such as a brando:GovernanceModel or governance-decision node

These properties have brando:ImpactAssessment in their domain and allow you to join the assessment to the wider Brand Brain governance graph and, via external IDs, to enterprise AI governance registers.

Key properties (domain includes `brando:ImpactAssessment`)

Below are the main properties intended to be used with brando:ImpactAssessment. They can also be reused on other governance classes where appropriate, but are described here from the impact assessment perspective.

1. Scope and subject of assessment

Property IRI	Range	Type	Comment (summary)
`brando:assessesBrand`	`brando:Brand`	`owl:ObjectProperty`	Brand whose reputation, trust, or equity is impacted by this AI system or use case.
`brando:assessesContext`	`brando:Context`	`owl:ObjectProperty`	Specific context (channel, audience, region) in which the AI behaviour is evaluated.
`brando:assessesModel`	`brando:BrandModel`	`owl:ObjectProperty`	Brand model, agent, or AI system that is the primary subject of this assessment.
`brando:assessesPrompt`	`brando:Prompt`	`owl:ObjectProperty`	Prompt, workflow, or orchestration pattern being assessed (e.g. a particular runbook or flow).
`brando:assessesAutomationRule`	`brando:AutomationRule`	`owl:ObjectProperty`	Automation or routing rule whose impact is being assessed (e.g. auto-send, auto-escalate).
`brando:assessmentScope`	`xsd:string`	`owl:DatatypeProperty`	Free-text or JSON-encoded description of scope (use case boundary, channels, jurisdictions, etc.).
`brando:assessmentUseCaseId`	`xsd:string`	`owl:DatatypeProperty`	Stable identifier for the use case in internal inventories or registers.
`brando:assessmentPhase`	`xsd:string`	`owl:DatatypeProperty`	Stage of the lifecycle (e.g. “design”, “pilot”, “production”, “retrospective”).

Usage notes (non-normative):

assessesBrand, assessesContext, and assessesModel are the primary anchors used for queries like:
“show all approved AI use cases for Brand X in Region Y”
“list impact assessments for this model.”
assessmentScope is deliberately flexible. Common patterns:
simple string: "Retail wealth advisory chatbot for UK and EU", or

JSON string:

{
  "channels": ["mobile app", "web"],
  "jurisdictions": ["UK", "DE"],
  "audiences": ["retail-high-net-worth"]
}

assessmentUseCaseId is useful for connecting the graph to:
internal AI inventories / registers,
risk registers,
internal LLM Suite registries or AI catalogs.

Example:

{
  "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial",
  "@type": "brando:ImpactAssessment",
  "brando:assessesBrand": { "@id": "https://example.com/brand/northstar" },
  "brando:assessesContext": { "@id": "https://example.com/brand/northstar/context/wealth-chat" },
  "brando:assessesModel": { "@id": "https://example.com/brand/northstar/models/wealth-assistant-v1" },
  "brando:assessmentUseCaseId": "NSB-WEALTH-CHAT-001",
  "brando:assessmentPhase": "design"
}

2. Risks, impacts, and harms

Property IRI	Range	Type	Comment (summary)
`brando:impactSummary`	`xsd:string`	`owl:DatatypeProperty`	Short human-readable summary of the overall brand, customer, and regulatory-signal impact.
`brando:impactAreas`	`xsd:string`	`owl:DatatypeProperty`	List or JSON string describing impacted dimensions (e.g. “brand trust”, “vulnerable customers”, “ESG”).
`brando:harmScenarios`	`xsd:string`	`owl:DatatypeProperty`	Example scenarios where the use case could cause harm or unacceptable outcomes.
`brando:riskRating`	`xsd:string`	`owl:DatatypeProperty`	Overall brand/impact rating for the use case (e.g. “Low”, “Medium”, “High”, “VeryHigh”).
`brando:residualRiskRating`	`xsd:string`	`owl:DatatypeProperty`	Risk rating after mitigations and controls are applied.
`brando:riskFactors`	`xsd:string`	`owl:DatatypeProperty`	Key factors that drive risk up or down (e.g. audience sensitivity, autonomy level, data sensitivity).
`brando:affectedPersonas`	`xsd:string`	`owl:DatatypeProperty`	Product/synthetic personas most affected by this use case (free text or references encoded as strings).

Usage notes (non-normative):

impactAreas, harmScenarios, and riskFactors work well as JSON arrays stored as strings, e.g.:

[
  "Potential mis-selling of complex financial products",
  "Over-reliance on AI responses during market stress",
  "Disproportionate impact on vulnerable or less literate customers"
]

riskRating and residualRiskRating can be aligned to:
internal brand/AI risk scales (e.g. L, M, H), or
a 1–5 style scale if harmonised with a wider risk vocabulary.

Example:

{
  "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial",
  "@type": "brando:ImpactAssessment",
  "brando:impactSummary": "High customer-impact, high brand-sensitivity use case for retail wealth advice. Medium residual risk with mitigations.",
  "brando:impactAreas": [
    "brand-trust",
    "customer-outcomes",
    "regulatory-exposure"
  ],
  "brando:harmScenarios": [
    "Customer interprets general guidance as personalised investment advice.",
    "Assistant encourages unsuitable risk profiles during market volatility."
  ],
  "brando:riskRating": "High",
  "brando:residualRiskRating": "Medium",
  "brando:riskFactors": [
    "Retail, non-professional investors",
    "Complex financial instruments",
    "High reliance on natural-language explanations"
  ]
}

3. Mitigations, controls, and decisions

Property IRI	Range	Type	Comment (summary)
`brando:linkedPolicy`	`brando:Policy`	`owl:ObjectProperty`	Policies that govern this use case (brand, legal, conduct, regional, product-specific).
`brando:requiredMitigation`	`xsd:string`	`owl:DatatypeProperty`	Mitigations or controls required before or during deployment (guard rails, escalation, human-in-the-loop).
`brando:mitigationStatus`	`xsd:string`	`owl:DatatypeProperty`	Status of mitigations (e.g. “planned”, “in-progress”, “implemented”, “not-feasible”).
`brando:decisionStatus`	`xsd:string`	`owl:DatatypeProperty`	Governance decision outcome for the use case (e.g. “approved”, “approved-with-conditions”, “rejected”, “pilot”).
`brando:decisionRationale`	`xsd:string`	`owl:DatatypeProperty`	Human-readable rationale explaining why the decision was taken, trade-offs, and assumptions.
`brando:reviewedByRole`	`schema:Role`	`owl:ObjectProperty`	Role (team or governance function) responsible for the assessment and decision (e.g. Brand, Legal, Risk).

Usage notes (non-normative):

linkedPolicy is the glue between impact assessments and machine-readable policies (brando:Policy) that Brand Brain runtimes actually enforce.
requiredMitigation is often easiest as a list of bullet-style strings; it can also hold JSON describing more structured actions.
decisionStatus is free-form but benefits from a constrained vocabulary in practice, e.g.:
"approved", "approved-with-conditions", "pilot-only", "rejected", "needs-more-information".

Example:

{
  "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial",
  "@type": "brando:ImpactAssessment",
  "brando:linkedPolicy": [
    { "@id": "https://example.com/brand/northstar/policies/global-brand-safety" },
    { "@id": "https://example.com/brand/northstar/policies/regulated-claims-uk" }
  ],
  "brando:requiredMitigation": [
    "Enforce refusal strategies for personalised investment advice.",
    "Route high-risk conversations to human advisors after 2 follow-up questions.",
    "Log and monitor conversations mentioning vulnerable customers for QA review."
  ],
  "brando:mitigationStatus": "in-progress",
  "brando:decisionStatus": "approved-with-conditions",
  "brando:decisionRationale": "High potential customer impact; residual risk acceptable with strict guard rails and human escalation."
}

4. Lifecycle, timing, and alignment with standards

Property IRI	Range	Type	Comment (summary)
`brando:assessmentDate`	`xsd:date`	`owl:DatatypeProperty`	Date when the core impact assessment was completed or approved.
`brando:nextReviewDue`	`xsd:date`	`owl:DatatypeProperty`	Date by which the assessment should be reviewed or renewed.
`brando:assessmentOwnerRole`	`schema:Role`	`owl:ObjectProperty`	Role accountable for maintaining this assessment over time (e.g. “Head of Brand Strategy”).
`brando:assessmentMethod`	`xsd:string`	`owl:DatatypeProperty`	Short description of the method/framework used (e.g. “ISO/IEC 42005-style brand impact lens on AI use case”).
`brando:externalSystemId`	`xsd:string`	`owl:DatatypeProperty`	Identifier for this assessment in external governance tools or AI registries (enterprise AI/ML risk systems).

Usage notes (non-normative):

assessmentMethod is where you can make standards alignment explicit, e.g.:

"assessmentMethod": "Aligned with ISO/IEC 42005 structure; scoped to brand, customer, and communications impact."

nextReviewDue supports operational hygiene:
dashboards can show which assessments are “stale”,
agents can automatically flag use cases due for re-assessment (e.g. after a model or policy change).
externalSystemId is a natural bridge into:
enterprise AI registers,
model risk systems,
DPIA / PIA registers (where applicable).

Example:

{
  "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial",
  "@type": "brando:ImpactAssessment",
  "brando:assessmentDate": "2025-03-15",
  "brando:nextReviewDue": "2025-09-15",
  "brando:assessmentMethod": "ISO/IEC 42005-aligned AI brand impact assessment template layered on top of enterprise AI/ML risk review.",
  "brando:externalSystemId": "AIREG-NSB-001"
}

5. Value, outcomes, and measurement

Property IRI	Range	Type	Comment (summary)
`brando:successCriteria`	`xsd:string`	`owl:DatatypeProperty`	Definition of success in brand, CX, or commercial terms (e.g. “reduce review time by 30% while maintaining NPS”).
`brando:expectedOutcomes`	`xsd:string`	`owl:DatatypeProperty`	Expected qualitative and quantitative outcomes versus baseline.
`brando:measurementPlan`	`xsd:string`	`owl:DatatypeProperty`	Description of how success will be measured (metrics, data sources, cadence, owners).
`brando:keyMetrics`	`xsd:string`	`owl:DatatypeProperty`	List or JSON string of key metrics tracked for this use case (e.g. AHT, brand consistency score, CSAT, revenue).
`brando:baselineSnapshot`	`xsd:string`	`owl:DatatypeProperty`	Baseline metrics or conditions before deployment (typically a JSON string or structured summary).
`brando:observedOutcomes`	`xsd:string`	`owl:DatatypeProperty`	Observed qualitative and quantitative impact after deployment / review cycles.
`brando:telemetrySource`	`schema:SoftwareApplication`	`owl:ObjectProperty`	Analytics or BI systems feeding measurements into this assessment (e.g. call centre BI, CMS logs, web analytics).

Usage notes (non-normative):

These properties turn brando:ImpactAssessment into a closed loop: not just “is it safe?” but “is it working and how well?”.
successCriteria should read like a contract between brand, product, and governance, e.g.:

Automate 25% of routine wealth queries while maintaining or improving brand trust and complaint rates.

keyMetrics, baselineSnapshot, and observedOutcomes typically contain JSON strings, e.g.:

{
  "avg_handling_time": 520,
  "complaint_rate_mis-selling": 0.8,
  "brand_consistency_score": 72
}

telemetrySource helps wire the assessment to real data platforms, so dashboards and agents can automatically pull metrics.

Example:

{
  "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial",
  "@type": "brando:ImpactAssessment",
  "brando:successCriteria": [
    "Automate at least 25% of routine wealth queries.",
    "Maintain or improve brand trust and mis-selling complaint rates vs baseline."
  ],
  "brando:expectedOutcomes": [
    "20–30% reduction in average handling time for routine wealth queries.",
    "Improved perceived clarity of communications in post-interaction surveys."
  ],
  "brando:measurementPlan": "Compare AHT, escalation rates, complaint rates, CSAT, and brand consistency for wealth chat before/after rollout over 90 days.",
  "brando:keyMetrics": [
    "avg_handling_time",
    "complaint_rate_mis-selling",
    "wealth_chat_csat",
    "brand_consistency_score"
  ],
  "brando:baselineSnapshot": "{\"avg_handling_time\":520,\"complaint_rate_mis-selling\":0.8,\"brand_consistency_score\":72}",
  "brando:observedOutcomes": "{\"avg_handling_time\":410,\"complaint_rate_mis-selling\":0.7,\"brand_consistency_score\":79}",
  "brando:telemetrySource": {
    "@id": "https://example.com/systems/contact-centre-bi",
    "@type": "schema:SoftwareApplication",
    "schema:name": "Contact Centre BI"
  }
}

6. Files and documentation

As with brando:Policy, these shared properties are often used to attach artefacts to impact assessments:

Property IRI	Range	Type	Comment (summary)
`brando:downloadableFile`	`schema:MediaObject`	`owl:ObjectProperty`	Links an impact assessment to the current canonical assessment document (e.g. PDF, doc, structured template export).
`brando:supportingFile`	`schema:MediaObject`	`owl:ObjectProperty`	Links to supporting materials such as risk analyses, user research, or design artefacts.

Example:

{
  "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial",
  "@type": "brando:ImpactAssessment",
  "brando:downloadableFile": {
    "@id": "https://assets.example.com/northstar/impact/wealth-assistant-v1-initial.pdf",
    "@type": "schema:MediaObject",
    "schema:name": "AI Brand Impact & Value Assessment – Wealth Assistant v1 (PDF)"
  },
  "brando:supportingFile": [
    {
      "@id": "https://assets.example.com/northstar/research/wealth-persona-research-2025.pdf",
      "@type": "schema:MediaObject",
      "schema:name": "Wealth Customer Persona Research 2025"
    }
  ]
}

Example: Brand + Context + Model + Impact Assessment

A combined example showing how brando:ImpactAssessment connects the Brand Brain governance objects, coexists with enterprise AI governance, and tracks value over time:

{
  "@context": {
    "schema": "https://schema.org/",
    "brando": "https://brandoschema.com/"
  },
  "@graph": [
    {
      "@id": "https://example.com/brand/northstar",
      "@type": "brando:Brand",
      "schema:name": "Northstar Bank",
      "brando:hasContext": {
        "@id": "https://example.com/brand/northstar/context/wealth-chat"
      },
      "brando:hasPolicy": {
        "@id": "https://example.com/brand/northstar/policies/global-brand-safety"
      }
    },
    {
      "@id": "https://example.com/brand/northstar/context/wealth-chat",
      "@type": "brando:Context",
      "brando:audienceSegment": "Retail wealth customers",
      "brando:domainContext": {
        "channel": "chat",
        "surface": "mobile-app",
        "region": "UK"
      },
      "brando:hasPolicy": {
        "@id": "https://example.com/brand/northstar/policies/regulated-claims-uk"
      }
    },
    {
      "@id": "https://example.com/brand/northstar/models/wealth-assistant-v1",
      "@type": "brando:BrandModel",
      "schema:name": "Wealth Assistant v1",
      "brando:hasImpactAssessment": {
        "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial"
      }
    },
    {
      "@id": "https://example.com/brand/northstar/assessments/wealth-assistant-v1-initial",
      "@type": "brando:ImpactAssessment",
      "brando:assessesBrand": { "@id": "https://example.com/brand/northstar" },
      "brando:assessesContext": { "@id": "https://example.com/brand/northstar/context/wealth-chat" },
      "brando:assessesModel": { "@id": "https://example.com/brand/northstar/models/wealth-assistant-v1" },
      "brando:assessmentUseCaseId": "NSB-WEALTH-CHAT-001",
      "brando:assessmentPhase": "design",
      "brando:impactSummary": "High-impact wealth advisory chatbot; approved with strict brand and conduct guard rails, human escalation, and clear success metrics.",
      "brando:impactAreas": [
        "brand-trust",
        "customer-outcomes",
        "regulatory-exposure"
      ],
      "brando:harmScenarios": [
        "Customer interprets generic explanations as personalised advice.",
        "Assistant downplays risks in volatile markets."
      ],
      "brando:riskRating": "High",
      "brando:residualRiskRating": "Medium",
      "brando:linkedPolicy": [
        { "@id": "https://example.com/brand/northstar/policies/global-brand-safety" },
        { "@id": "https://example.com/brand/northstar/policies/regulated-claims-uk" }
      ],
      "brando:requiredMitigation": [
        "Hard refusal on personalised investment advice.",
        "Escalate to human advisor for complex or vulnerable-customer situations.",
        "Monitor conversation samples weekly for mis-selling risks."
      ],
      "brando:decisionStatus": "approved-with-conditions",
      "brando:decisionRationale": "Residual brand and customer risk acceptable with guard rails, human-in-the-loop oversight, and active monitoring of success metrics.",
      "brando:assessmentMethod": "ISO/IEC 42005-aligned AI brand impact assessment layered on top of enterprise AI model risk review.",
      "brando:assessmentDate": "2025-03-15",
      "brando:nextReviewDue": "2025-09-15",
      "brando:externalSystemId": "AIREG-NSB-001",
      "brando:successCriteria": [
        "Automate at least 25% of routine wealth queries.",
        "Maintain or improve wealth-chat CSAT and brand trust scores.",
        "No increase in mis-selling or suitability complaint rates."
      ],
      "brando:expectedOutcomes": [
        "20–30% reduction in AHT for routine wealth queries.",
        "Improved perceived clarity and reassurance during market volatility."
      ],
      "brando:measurementPlan": "Track AHT, escalation rate, complaint rate, CSAT, and brand consistency for wealth chat in monthly governance reviews over the first 90 days.",
      "brando:keyMetrics": [
        "avg_handling_time",
        "escalation_rate",
        "complaint_rate_mis-selling",
        "wealth_chat_csat",
        "brand_consistency_score"
      ],
      "brando:baselineSnapshot": "{\"avg_handling_time\":520,\"complaint_rate_mis-selling\":0.8,\"wealth_chat_csat\":4.2,\"brand_consistency_score\":72}",
      "brando:observedOutcomes": "{\"avg_handling_time\":410,\"complaint_rate_mis-selling\":0.7,\"wealth_chat_csat\":4.5,\"brand_consistency_score\":79}",
      "brando:telemetrySource": {
        "@id": "https://example.com/systems/contact-centre-bi",
        "@type": "schema:SoftwareApplication",
        "schema:name": "Contact Centre BI"
      },
      "brando:retrievableInLLM": true
    }
  ]
}

Runtime pattern (non-normative):

Resolve current brando:Brand, brando:Context, and brando:BrandModel.
Lookup associated brando:ImpactAssessment nodes.
Filter to:
decisionStatus ∈ approved, approved-with-conditions
assessmentDate and nextReviewDue still valid.
Use linkedPolicy to pull the precise brando:Policy guard rails.
Use successCriteria, keyMetrics, baselineSnapshot, and observedOutcomes to:
populate dashboards for value tracking,
drive alerts when value drifts or metrics degrade,
inform re-assessments.
Use externalSystemId / assessmentMethod to show how this use case maps into wider enterprise AI governance.

Concepts / overview
- Brando types overview
- Getting started
Related types
Properties index
- Properties overview

brando:ImpactAssessment

Term definition

Intended usage

Relationships

Inbound

Key properties (domain includes brando:ImpactAssessment)

1. Scope and subject of assessment

2. Risks, impacts, and harms

3. Mitigations, controls, and decisions

4. Lifecycle, timing, and alignment with standards

5. Value, outcomes, and measurement

6. Files and documentation

Example: Brand + Context + Model + Impact Assessment

Related documentation

`brando:ImpactAssessment`

Key properties (domain includes `brando:ImpactAssessment`)