unified
UC-AI-04 — Assess impacts and classify AI systems before deployment
Operate a documented AI impact assessment process performed before deployment and repeated on significant change, assessing consequences for individuals, groups of individuals, and society, including fairness, safety, and fundamental-rights effects. Classify each system against applicable regulatory risk categories, including any high-risk designation, and record the classification rationale. Retain assessment reports, approvals, and reassessment triggers as evidence.
Record JSON · Open in map · Data retrieval guide
Catalog revision: 24028ffcfc2b295fa1b08ee6caa84b765f0731b321496bf4f548c49ad2177028. A connection does not establish full coverage.
Attributes
- domain
- AI Governance
- type
- preventive
- category
- administrative
Details
- unified_id
- UC-AI-04
- title
- Assess impacts and classify AI systems before deployment
- statement
- Operate a documented AI impact assessment process performed before deployment and repeated on significant change, assessing consequences for individuals, groups of individuals, and society, including fairness, safety, and fundamental-rights effects. Classify each system against applicable regulatory risk categories, including any high-risk designation, and record the classification rationale. Retain assessment reports, approvals, and reassessment triggers as evidence.
- domain
- AI Governance
- control_type
- preventive
- control_category
- administrative
- members
- framework
- iso-42001
- control_id
- A.5.2
- coverage
- full
- relationship
- superset_of
- framework
- iso-42001
- control_id
- A.5.3
- coverage
- full
- relationship
- superset_of
- framework
- iso-42001
- control_id
- A.5.4
- coverage
- full
- relationship
- superset_of
- framework
- iso-42001
- control_id
- A.5.5
- coverage
- full
- relationship
- superset_of
- framework
- eu-ai-act
- control_id
- AIA-Art6-7
- coverage
- full
- relationship
- superset_of
- framework
- aiuc-1
- control_id
- C001
- coverage
- partial
- delta
- a system-specific AI risk taxonomy with severity tiers that drives output filtering, monitoring categories, and third-party test scope
- relationship
- intersects_with
- framework
- eu-ai-act
- control_id
- AIA-Art27
- coverage
- full
- relationship
- superset_of
- guidance
Source
No record-specific source URL is provided.
Connections
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Unfair exclusion by AI in education and training
- strength
- primary
- rationale
- Pre-deployment impact assessment plus high-risk classification is the mandated control education-AI is faulted for lacking, gating unfair scoring/exclusion.
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Rights violations from AI in law enforcement
- strength
- primary
- rationale
- Fundamental-rights impact assessment and high-risk classification is the safeguard law-enforcement AI is faulted for lacking before deployment.
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Discriminatory outcomes from AI in employment
- strength
- primary
- rationale
- Impact/DPIA assessment and classification gates biased or opaque employment-AI before deployment, the assessment the risk says is missing.
- ISO/IEC 42001 AI Management System Internal Audit tests UC-AI-04 — Assess impacts and classify AI systems before deployment
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Public-safety harm from AI in critical infrastructure
- strength
- primary
- rationale
- Pre-deployment safety-impact assessment and risk classification is the risk-management step critical-infrastructure AI is faulted for operating without.
- UC-AI-04 — Assess impacts and classify AI systems before deployment maps_to A.5.2 — AI system impact assessment process
- framework
- iso-42001
- control_id
- A.5.2
- coverage
- full
- relationship
- superset_of
- delta
- Not provided
- source_version
- 2023
- provenance
- mapper
- coworkcanvas-compliance-graph
- reviewDate
- 2026-09-07
- direction
- canonical_to_source
- defaultConfidence
- medium
- defaultStatus
- active
- note
- Each member is a documented relationship claim from the canonical unified control to a source control or guidance proposition. relationship: equal|superset_of (full) / intersects_with|subset_of (partial) / informs (guidance). confidence 'medium' = single-mapper, documented, not yet externally corroborated. source_version is the member framework's edition from the standard version register.
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Wrongful denial by AI in migration and border control
- strength
- primary
- rationale
- Pre-deployment impact assessment and rights-safeguard classification gate the wrongful migration/border decisions the risk says lack them.
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Harm to due process and democratic integrity from AI
- strength
- primary
- rationale
- Impact assessment of effects on individuals/groups/society and fundamental rights is the core high-risk control gating justice/democracy harms before deployment.
- UC-AI-04 — Assess impacts and classify AI systems before deployment maps_to A.5.4 — Assessing AI system impact on individuals or groups of individuals
- framework
- iso-42001
- control_id
- A.5.4
- coverage
- full
- relationship
- superset_of
- delta
- Not provided
- source_version
- 2023
- provenance
- mapper
- coworkcanvas-compliance-graph
- reviewDate
- 2026-09-07
- direction
- canonical_to_source
- defaultConfidence
- medium
- defaultStatus
- active
- note
- Each member is a documented relationship claim from the canonical unified control to a source control or guidance proposition. relationship: equal|superset_of (full) / intersects_with|subset_of (partial) / informs (guidance). confidence 'medium' = single-mapper, documented, not yet externally corroborated. source_version is the member framework's edition from the standard version register.
- AI System Development, Data & Deployment Gate operates UC-AI-04 — Assess impacts and classify AI systems before deployment
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Rights harm from biometric-identification AI
- strength
- primary
- rationale
- Fundamental-rights impact assessment and classification is the exact control biometric-ID systems are faulted for lacking, blocking unassessed deployment.
- UC-AI-04 — Assess impacts and classify AI systems before deployment maps_to A.5.3 — Documentation of AI system impact assessments
- framework
- iso-42001
- control_id
- A.5.3
- coverage
- full
- relationship
- superset_of
- delta
- Not provided
- source_version
- 2023
- provenance
- mapper
- coworkcanvas-compliance-graph
- reviewDate
- 2026-09-07
- direction
- canonical_to_source
- defaultConfidence
- medium
- defaultStatus
- active
- note
- Each member is a documented relationship claim from the canonical unified control to a source control or guidance proposition. relationship: equal|superset_of (full) / intersects_with|subset_of (partial) / informs (guidance). confidence 'medium' = single-mapper, documented, not yet externally corroborated. source_version is the member framework's edition from the standard version register.
- UC-AI-04 — Assess impacts and classify AI systems before deployment maps_to C001 — Define AI risk taxonomy
- framework
- aiuc-1
- control_id
- C001
- coverage
- partial
- delta
- a system-specific AI risk taxonomy with severity tiers that drives output filtering, monitoring categories, and third-party test scope
- relationship
- intersects_with
- source_version
- July 15, 2026 release (quarterly update cadence)
- provenance
- mapper
- coworkcanvas-compliance-graph
- reviewDate
- 2026-09-07
- direction
- canonical_to_source
- defaultConfidence
- medium
- defaultStatus
- active
- note
- Each member is a documented relationship claim from the canonical unified control to a source control or guidance proposition. relationship: equal|superset_of (full) / intersects_with|subset_of (partial) / informs (guidance). confidence 'medium' = single-mapper, documented, not yet externally corroborated. source_version is the member framework's edition from the standard version register.
- AI Governance & Risk/Impact Assessment oversees UC-AI-04 — Assess impacts and classify AI systems before deployment
- UC-AI-04 — Assess impacts and classify AI systems before deployment maps_to A.5.5 — Assessing societal impacts of AI systems
- framework
- iso-42001
- control_id
- A.5.5
- coverage
- full
- relationship
- superset_of
- delta
- Not provided
- source_version
- 2023
- provenance
- mapper
- coworkcanvas-compliance-graph
- reviewDate
- 2026-09-07
- direction
- canonical_to_source
- defaultConfidence
- medium
- defaultStatus
- active
- note
- Each member is a documented relationship claim from the canonical unified control to a source control or guidance proposition. relationship: equal|superset_of (full) / intersects_with|subset_of (partial) / informs (guidance). confidence 'medium' = single-mapper, documented, not yet externally corroborated. source_version is the member framework's edition from the standard version register.
- UC-AI-04 — Assess impacts and classify AI systems before deployment maps_to AIA-Art27 — Fundamental rights impact assessment for high-risk AI systems (deployers)
- framework
- eu-ai-act
- control_id
- AIA-Art27
- coverage
- full
- relationship
- superset_of
- delta
- Not provided
- source_version
- Regulation (EU) 2024/1689
- provenance
- mapper
- coworkcanvas-compliance-graph
- reviewDate
- 2026-09-07
- direction
- canonical_to_source
- defaultConfidence
- medium
- defaultStatus
- active
- note
- Each member is a documented relationship claim from the canonical unified control to a source control or guidance proposition. relationship: equal|superset_of (full) / intersects_with|subset_of (partial) / informs (guidance). confidence 'medium' = single-mapper, documented, not yet externally corroborated. source_version is the member framework's edition from the standard version register.
- UC-AI-04 — Assess impacts and classify AI systems before deployment maps_to AIA-Art6-7 — Risk-based classification of high-risk AI systems
- framework
- eu-ai-act
- control_id
- AIA-Art6-7
- coverage
- full
- relationship
- superset_of
- delta
- Not provided
- source_version
- Regulation (EU) 2024/1689
- provenance
- mapper
- coworkcanvas-compliance-graph
- reviewDate
- 2026-09-07
- direction
- canonical_to_source
- defaultConfidence
- medium
- defaultStatus
- active
- note
- Each member is a documented relationship claim from the canonical unified control to a source control or guidance proposition. relationship: equal|superset_of (full) / intersects_with|subset_of (partial) / informs (guidance). confidence 'medium' = single-mapper, documented, not yet externally corroborated. source_version is the member framework's edition from the standard version register.
- UC-AI-04 — Assess impacts and classify AI systems before deployment mitigates Unlawful denial of essential services by AI
- strength
- primary
- rationale
- Impact assessment plus conformity classification gates unfair benefit/credit/insurance decisions the risk says deploy without required controls.