ipfour
AI ServicesAI GovernanceAI Impact Assessment
AI Governance

Formal impact assessment for AI decision systems.

When AI makes or influences decisions affecting people, a formal impact assessment is not optional. We produce assessments aligned with ICO requirements and UK AI regulation guidance, giving you documented evidence of responsible AI deployment.

ICO Aligned
UK AI Act Ready
Decision Systems
Harm Assessment
4-6 weeks
Typical impact assessment delivery time
ICO Aligned
Assessments satisfying ICO requirements
UK AI Act
Readiness documentation included
UK-Wide
Impact assessments delivered across sectors
What We Assess

Every dimension of AI decision-making impact.

A complete AI impact assessment covers the system, the people it affects, the harms it could cause, and the safeguards that mitigate those harms.

AI System Scoping and Description

A clear description of the AI system being assessed, including its purpose, the decisions it makes or influences, the data it uses, and the people it affects.

System DescriptionDecision MappingScope Definition

Affected Persons Analysis

Identification and analysis of all groups of people affected by the AI system, including direct users, people subject to AI-influenced decisions, and third parties.

Affected GroupsStakeholder AnalysisVulnerable Groups

Proportionality and Necessity Assessment

Assessment of whether the use of AI is proportionate and necessary for the stated purpose, and whether less privacy-intrusive or less risky alternatives exist.

ProportionalityNecessity TestAlternatives Analysis

Harm and Risk Identification

Systematic identification of potential harms to individuals and groups from the AI system, including discrimination, financial harm, reputational harm, and loss of rights.

Harm AssessmentDiscrimination RiskRights Impact

Safeguards and Mitigations

Documentation of the safeguards in place to mitigate identified harms, including human oversight, appeal mechanisms, transparency measures, and monitoring arrangements.

SafeguardsAppeal RightsTransparency

UK AI Act Readiness Documentation

Assessment of the AI system against emerging UK AI regulation requirements, producing documentation that demonstrates compliance readiness as the regulatory landscape develops.

UK AI ActRegulatory ReadinessCompliance Documentation
How We Work

From system discovery to approved assessment.

01

AI System Discovery

We work with your technical and operational teams to understand exactly how the AI system works, what decisions it makes or influences, and what data it processes.

02

Stakeholder Identification

We identify all stakeholders affected by the AI system, including those who may be subject to AI-influenced decisions, and assess the potential impact on each group.

03

Harm and Risk Assessment

We systematically assess the potential harms from the AI system, rating each by likelihood and severity, and identifying the groups most at risk.

04

Safeguard Review

We review the safeguards currently in place and assess whether they are sufficient to mitigate the identified harms to an acceptable level.

05

Impact Assessment Drafting

We produce the formal impact assessment document, structured to satisfy ICO requirements and aligned with UK AI governance guidance from DSIT and the AI Safety Institute.

06

Review and Monitoring Schedule

We establish a schedule for reviewing and updating the impact assessment as the AI system evolves, regulations change, and new evidence of impact emerges.

Real Results

Impact assessments for regulated AI systems.

Local Authority Housing

A local authority used an AI system to prioritise housing applications. The council needed a formal impact assessment before the system could be approved for continued use.

Full AI impact assessment produced. Three safeguards strengthened following the assessment. Assessment approved by the council's data protection officer and legal team. System approved for continued use with enhanced monitoring.

Insurance Underwriter

An insurance firm used AI to assist underwriters in pricing decisions. The FCA had begun scrutinising algorithmic decision-making and the firm needed documented impact assessment.

Impact assessment produced covering pricing fairness, discrimination risk, and customer harm. FCA inquiry satisfied. Firm now has a repeatable process for assessing new AI-assisted underwriting tools.

NHS-Contracted Diagnostic Service

A diagnostic imaging service used AI to assist radiologists in identifying abnormalities. An NHS contract renewal required formal documentation of the AI system's impact assessment.

Impact assessment produced aligned with NHS AI governance requirements. Clinical oversight safeguards documented. Contract renewed. Assessment used as a template for two further AI diagnostic tools.

Get Started

Does your AI system need a formal impact assessment?

Book a free AI governance review. We will assess whether your AI systems require a formal impact assessment, what it needs to cover, and how quickly we can deliver it.