Questions about EIAs
and EqualityIQ

An equality impact assessment (EIA) is a structured analysis of how a proposed policy, service change, or decision is likely to affect people with different protected characteristics under the Equality Act 2010.

UK public authorities, including local councils, carry out EIAs as part of their Public Sector Equality Duty (PSED). A good EIA identifies both potential negative impacts and opportunities to advance equality, supported by evidence from research and local demographic data.

The nine protected characteristics under the Equality Act 2010 are:

• Age
• Disability
• Race
• Sex
• Religion or belief
• Pregnancy and maternity
• Marriage and civil partnership
• Sexual orientation
• Gender reassignment

A thorough EIA must consider the likely impact of a proposal on each of these groups, including intersecting impacts where someone may belong to more than one.

A thorough EIA should include:

• A description of the proposal and its objectives
• Identification of who is likely to be affected, by protected characteristic
• Evidence from research and local data (such as Census 2021)
• Assessment of whether impacts are positive, negative, or neutral
• Proposed mitigating actions where negative impacts are identified
• A monitoring and review plan
• Details of any consultation carried out with affected groups

For a straightforward proposal, a thorough EIA typically takes an experienced equality officer one to two days to research and draft — longer for complex or high-impact decisions. Most of that time is spent on research and evidence-gathering rather than writing.

That research burden is why thorough assessment tends to be rationed to the most visible decisions. EqualityIQ produces the evidenced first draft for every assessment, so the same depth of analysis reaches the routine policy as the flagship one — your officers spend their time on judgement, local knowledge and consultation rather than literature searching.

Too often, yes — and when they are, courts and scrutiny committees notice: decisions have been quashed where the assessment was produced too late or asserted conclusions without evidence. The cause is rarely the officer. It's analysis commissioned days before a deadline, templates with no research behind them, and no time to do better. EqualityIQ changes that: the evidenced first draft arrives at the start, every claim is cited to a named source, and actions are tracked after sign-off — so the assessment shapes the decision rather than documenting it afterwards.

Yes. Under the Public Sector Equality Duty (Section 149 of the Equality Act 2010), public bodies must have "due regard" to equality when exercising their functions. While the law does not mandate a specific EIA format, documented equality analysis is the recognised way to demonstrate compliance.

Failure to demonstrate due regard has resulted in successful judicial reviews against local authorities — including cases where an EIA existed but was produced too late or lacked sufficient evidence.

The PSED is a legal requirement under Section 149 of the Equality Act 2010. It requires public bodies, including local councils, to have due regard to three aims:

• Eliminating discrimination, harassment, and victimisation
• Advancing equality of opportunity between those who share a protected characteristic and those who do not
• Fostering good relations between different groups

The duty applies to all nine protected characteristics and is ongoing — not a one-off compliance exercise at the point of a single decision.

Non-compliance can result in judicial review of the decision, which may lead to the decision being quashed and having to be retaken. Organisations can also face EHRC investigations, reputational damage, and the significant cost and disruption of legal proceedings.

Beyond legal risk, poorly evidenced equality analysis means decisions are more likely to produce unintended harms that require expensive remediation — often far exceeding the cost of doing the EIA properly in the first place.

A standard EIA template provides structure but no analytical content — the research, evidence-gathering, and analysis still falls entirely on the officer completing it.

EqualityIQ generates substantive, cited analysis for each protected characteristic, drawing on a curated knowledge base developed around the expertise of Laura Walker FIEDIP — Director of Equality Impact Consulting Ltd, with over 22 years' experience. The result is a well-evidenced first draft that an equality officer reviews, contextualises, and adapts — rather than starting from a blank page.

Every subscription includes a generous fair-usage allowance, set well above what even a busy equality team typically needs. You'll be notified well before you approach it, with the option to request an increase. Details are set out in our Terms of Service.

AI can play a valuable role in EIA work — particularly in synthesising research evidence, structuring analysis across all nine characteristics, and reducing the time burden on equality officers.

However, AI cannot replace officer judgment, local knowledge, or community consultation. EqualityIQ is designed as a tool to support equality officers, producing a well-researched first draft that officers review and sign off. Professional accountability for the final EIA remains with your organisation.

No — and it is not designed to. EqualityIQ gives your equality officers a well-researched, evidence-cited first draft. They review, contextualise with local knowledge, engage with affected communities, and sign off the final assessment.

The expertise stays with your team. The research burden is significantly reduced.

EqualityIQ is priced as an organisation-wide annual subscription, tailored to your organisation's size and assessment volume. One subscription covers your whole organisation, with no per-seat fees and no per-assessment charges.

See our pricing page or contact us and we'll come back with a straightforward quote for your organisation.

Yes. The Public Sector Equality Duty applies to all listed public authorities — including further education colleges — and housing associations carrying out public functions. EqualityIQ supports councils, colleges and housing associations, with the evidence base tailored to each sector: education and student-population data for colleges, resident and tenant demographics for councils and housing associations. The workflow, dashboards and expert certification options are the same across sectors.

EqualityIQ draws on a curated knowledge base of research from approved organisations relevant to each protected characteristic, including Age UK, Mencap, the Runnymede Trust, Scope, Stonewall, Barnardo's, the Fawcett Society, the NSPCC, and the Early Intervention Foundation.

Secondary data sources include Census 2021 (ONS) and OHID Fingertips for local population and health outcome data. All citations are transparent and traceable so officers can verify and follow up sources.

The mapping of organisations to protected characteristics, and the standard applied to each, was developed by Laura Walker FIEDIP based on 22 years of equality assessment practice.

The knowledge base is updated regularly as new research is published by approved organisations. New findings are reviewed against the same practitioner-set standards before entering the knowledge base, so assessments draw on current evidence rather than relying solely on older publications.

Where a specific, recent publication is relevant to a proposal, officers can also add supplementary evidence to their assessment during the review process.

Yes. Proposal data submitted through EqualityIQ is never used to train AI models and is handled in accordance with UK GDPR requirements. We can provide a Data Processing Agreement (DPA) for organisations that require one.

If you have specific data governance requirements, contact us to discuss how EqualityIQ can meet them.

EqualityIQ is hosted on UK-based infrastructure — our servers are located in Manchester — and your organisation's data is stored in the UK. Assessment generation is processed by Anthropic's Claude via its API under a data processing agreement: your data is never used to train AI models, and international transfers are protected by UK GDPR safeguards. Contact us for our full data residency and transfer documentation.

Your organisation owns its assessments and data. You can export any assessment to PDF or Word at any time, and if you end your subscription we'll provide a complete export of your data on request and delete it from our systems in line with our terms and UK GDPR.

EqualityIQ can be procured via direct award — annual subscription values typically fall within most public bodies' direct-award thresholds — and we're happy to work within your organisation's preferred procurement route. We can provide the documentation your procurement and legal teams require.

Contact us to start a conversation about your organisation's procurement process.

Yes. We offer a free demo where we run EqualityIQ on one of your organisation's live proposals, so you can see the quality of output before signing anything. Request a demo and we'll arrange it within a few working days — no commitment, no setup required.

A single subscription covers your whole organisation — all departments, directorates, and teams. There are no per-seat fees and no per-assessment charges.

Subscriptions can also include a dedicated account manager and custom knowledge base entries, depending on your organisation's needs.

This is the right question to ask of any AI tool used in professional practice. EqualityIQ is designed specifically to reduce this risk: rather than generating analysis from its general training data, it draws on a curated knowledge base of research from named, approved organisations. Every claim in the output is tied to a specific publication from a specific organisation.

That said, officer review remains essential. EqualityIQ produces a well-evidenced first draft — not a finished document. Your equality officer should review every assessment before it is finalised.

Yes — through a structured review. Officers add local context and evidence, confirm or override the impact level for each characteristic, and can mark characteristics not relevant. The AI-generated analysis itself is locked once the report is generated, preserving a clear audit trail of what the AI produced versus what your officers added.

For full free-form editing, export to Word (.docx) — ideal for internal review with track changes before sign-off.

Your subscription includes the user accounts your organisation needs — agreed at setup, with no per-seat fees. Contact us to discuss your team's requirements.

Every assessment lives on a dashboard showing its status, risk rating and review date. Mitigations become tracked actions with owners and deadlines, overdue actions are flagged automatically, and review dates align to your policy review calendar — so each assessment resurfaces when its policy does. You can also see coverage across the three aims of the Public Sector Equality Duty.

The knowledge base covers a broad range of policy and service areas relevant to the public sector — housing, social care, education, transport, leisure, planning, and more. For most proposals, relevant evidence will be available across the majority of protected characteristics.

Where evidence is sparse, EqualityIQ will indicate this rather than generating unsupported analysis. Officers can supplement the output with their own research — which is itself good practice under the PSED.

EqualityIQ incorporates Census 2021 data which provides local authority and ward-level breakdowns for most protected characteristics. This means the analysis can reflect that your local area has a particularly high proportion of older residents, a significant South Asian community, or a high rate of disability claimants — rather than relying solely on national averages.

For very specific local context — a particular community group, a known local issue, or recent consultation findings — officers should add this during the review and editing phase.

No tool can guarantee PSED compliance — and any that claims otherwise should be treated with scepticism. Compliance depends on how the EIA is used in the decision-making process, not just whether one exists.

What EqualityIQ does is significantly strengthen the evidence base of your assessments and ensure consistent, documented analysis across all nine characteristics. An EqualityIQ assessment, properly reviewed and used to inform the decision, puts your organisation in a much stronger position than a manually produced assessment under time pressure.

Yes, after officer review. The EqualityIQ output is formatted as a structured EIA and can be exported directly to PDF or Word for inclusion in committee and board reports. Most organisations will want their equality officer to review and sign off the assessment before it goes to committee.

For most organisations, setup takes less than a day — and always within the 10 working days our terms commit to. Once your subscription is confirmed, we create your account, set up your users, and walk you through the system. There is no software to install — EqualityIQ is fully web-based and works in any modern browser.

Yes — this is available as an optional extension to your subscription. You can add organisation-specific research, local needs assessments, consultation findings, and JSNA data to a private knowledge base.

Contact us to discuss how this would work for your organisation.

We build EqualityIQ to align with WCAG 2.1 AA, publish an accessibility statement, and treat accessibility as continuous work rather than a one-off audit — as you'd expect from an equality product. If you use assistive technology and encounter a barrier, we want to hear about it and we'll fix it.

EqualityIQ is built on Claude, Anthropic's AI model, which is well regarded for its accuracy, careful reasoning, and reduced tendency to fabricate information. Anthropic has a strong focus on AI safety and reliability, which aligns with the professional context in which EqualityIQ is used.

No. Your proposal descriptions and assessment data are never used to train AI models — by EqualityIQ or by our underlying AI provider. Each assessment is processed and then the data remains in your account, not in any training pipeline.

Edit it. The output is a first draft and you are not obliged to accept any part of it. If analysis for a particular characteristic seems off, the officer reviewing the assessment can amend or replace that section.

You'll get something that looks like an EIA — fluent, structured, confident. What you won't get: citations to verified research (general chatbots routinely invent sources), your local population data, a defined assessment methodology, an audit trail of what was generated and what your officers reviewed, or any tracking of actions afterwards. There's also a governance problem: typing unpublished proposal details into a consumer chatbot places your decision-making information in a third-party service outside your control. EqualityIQ generates assessments inside a governed platform — expert methodology, approved evidence, local data, a locked audit trail and tracked actions — with a data processing agreement behind it. The difference isn't the writing. It's everything that makes the writing defensible.

EqualityIQ is designed for the routine equality impact assessments that make up the majority of an organisation's equality work — service changes, policy reviews, budget decisions, and operational proposals.

However, for high-stakes proposals — particularly those involving compulsory purchase orders, Development Consent Orders, major infrastructure schemes, or decisions with a significant risk of legal challenge — AI-generated analysis alone may not be sufficient. These proposals benefit from specialist expert involvement alongside the EqualityIQ assessment.

Yes. Laura Walker FIEDIP has extensive experience in equality impact assessment for Compulsory Purchase Orders and Development Consent Orders, including representing councils in CPO processes.

Contact her at [email protected] or visit equalityimpact.co.uk.