Diversity Hiring Software: How UK Employers Expand the Candidate Pool Compliantly

Jurisdiction: This article covers the Equality Act 2010, which applies in England, Wales, and Scotland. Northern Ireland has separate equality legislation.

This article is for informational purposes only and does not constitute legal advice. Employment law and screening regulations vary by jurisdiction. Consult a qualified legal professional before making compliance decisions.

Diversity hiring starts long before an interview. The candidates a team considers are shaped by where it looks, and for many UK employers that still means job boards, agencies, and referrals from people already on the payroll. Each channel has value, and each reaches only part of the available talent.

In the CIPD's Resourcing and talent planning 2026 report, 46% of organisations that recruited in the past year hired a more diverse workforce. Another 22% reported challenges in hiring diverse talent.

What does diversity hiring software actually do?

Diversity hiring software helps employers reach and fairly assess a broader range of candidates. The best tools widen who is found and invited to apply, and keep selection focused on skills and job-related criteria.

Most tools in the category support one or more of these jobs:

Sourcing: searching more platforms and profiles, so the pool reflects the wider labour market

Job adverts: checking language so adverts appeal to a wider audience

Structured screening: applying the same job-related criteria to every candidate

Reducing irrelevant information: removing details such as names from early screening, which 19% of organisations in the CIPD report already do

Monitoring: collecting voluntary diversity data directly from candidates to track outcomes

In summary, diversity hiring software works by widening reach and standardising assessment, not by selecting people because of who they are.

What does the Equality Act 2010 mean for sourcing, as opposed to selection?

The Equality Act 2010 allows employers to take proportionate steps to encourage applications from under-represented groups. The Act does not allow candidates to be selected because of a protected characteristic, apart from a narrow tie-breaker. That makes sourcing the natural place to focus diversity efforts.

The nine protected characteristics

The Act protects people from discrimination because of age, disability, gender reassignment, marriage and civil partnership, pregnancy and maternity, race, religion or belief, sex, and sexual orientation.

Positive action at the sourcing stage

Section 158 permits proportionate action where an employer reasonably thinks people who share a protected characteristic are disadvantaged or under-represented. At the sourcing stage, that can include:

Advertising roles in places that reach under-represented groups

Encouraging applications from groups with low participation

Widening sourcing channels beyond existing networks

The limits at selection

Section 159 allows an employer to prefer a candidate with a protected characteristic in limited circumstances. The candidate must be as qualified as the other, the employer must have no general policy of preferential treatment, and the action must be proportionate. Outside that tie-breaker, selection must rest on job-related criteria.

Here is how the two stages compare directly.

Stage

What employers can do

What employers cannot do

Sourcing and attraction

Widen channels, target outreach to under-represented groups, encourage applications

Exclude or deter candidates because of a protected characteristic

Screening and selection

Apply consistent, job-related criteria and structured assessment

Rank, filter, or select candidates by protected characteristic

Tie-breaker

Prefer an equally qualified candidate from an under-represented group, proportionately

Operate a general policy of automatic preference

In summary, the law supports widening the pool and expects selection to stay on merit. TraqCheck's guide to what it means to be an equal opportunity employer covers the wider principles.

Why does candidate pool breadth matter for diversity outcomes?

Pool breadth matters because selection can only be as diverse as the pool it draws from. If sourcing reaches the same networks every time, consistent and fair selection will still produce a narrow set of hires.

Referral and network hiring illustrates the point. The CIPD report found that 28% of private sector employers counted employee referral schemes among their most effective attraction methods, compared with 12% of public and non-profit organisations. Referrals can bring strong candidates, and they naturally draw on the networks of existing staff.

Broadening the pool brings three benefits:

More candidates with the right skills, including those outside existing networks

Better data on whether outreach is reaching under-represented groups

A stronger base for fair, structured selection

Measuring whether outreach is working

Breadth is only useful if teams can see its effect. Good practice keeps measurement separate from selection:

Collect diversity information voluntarily and directly from candidates, never by inference

Store it securely, away from the people making hiring decisions

Compare the make-up of the pool at each stage, from sourcing to offer

Report trends using anonymised statistics

The ICO's draft recruitment guidance gives a similar example of an employer collecting equality data through a secure portal and using it only to produce anonymised statistics. That approach helps talent acquisition teams see which channels widen the pool without the data ever influencing an individual decision.

How does AI sourcing expand reach beyond referral networks?

AI sourcing expands reach by searching far more platforms and profiles than a recruiter could cover by hand, and by matching on skills rather than familiar titles or employers. The result is qualified people who would not appear through referrals or a single job board.

NINA, TraqCheck's AI hiring agent, takes a role brief through simple chat and searches 50+ platforms in real time, drawing on 800M+ candidate profiles. NINA ranks matches on skills, intent, and availability rather than keywords alone, sends personalised outreach, and shortlists candidates who respond with genuine interest.

For diversity hiring, that breadth helps in three ways:

Reach: candidates come from many platforms rather than a single network

Skills focus: matching on skills and intent supports consistent, job-related criteria

Consistency: every candidate for a role is approached and assessed against the same brief

Your team reviews every shortlist and makes every decision, so selection stays with people and on merit.

What should you look for in diversity-minded sourcing tools?

Look for sourcing tools that widen reach, assess everyone against the same job-related criteria, and never filter or infer protected characteristics. The ICO's audits of AI recruitment tools give useful guidance on what good looks like.

The ICO's 2024 audit outcomes found that some tools allowed recruiters to filter out candidates with certain protected characteristics, and others inferred characteristics such as gender and ethnicity from names. The ICO instructed providers to collect accurate information directly from candidates and to monitor for discrimination.

Use that as a checklist when evaluating tools:

Breadth: how many platforms and profiles does the tool search?

Job-related matching: does it rank on skills and requirements set by your team?

No characteristic filters: can anyone filter or rank candidates by a protected characteristic? The answer should be no.

No inference: does the tool infer gender, ethnicity, or other characteristics? A well-designed tool does not.

Voluntary monitoring: is diversity data collected directly from candidates, separately from selection?

Human review: does a person review every shortlist and decision?

Transparency: can the provider explain how candidates are surfaced?

Building a broader pipeline

Diversity hiring works best when it starts at sourcing. Widen where you look, keep selection on consistent job-related criteria, and measure whether outreach is reaching the people it is meant to reach. The Equality Act gives employers clear room to broaden the pool, and AI sourcing makes that breadth practical at scale. Founders and lean teams can try NINA for free with one active role. Larger talent teams can book a demo to see how NINA widens sourcing for their open roles.

Frequently asked questions

How does diversity hiring software work?

Diversity hiring software widens the candidate pool through broader sourcing and more inclusive adverts, and supports consistent, job-related screening. Good tools never filter or rank candidates by protected characteristics.

Can AI sourcing improve diversity hiring?

Yes, by widening reach. AI sourcing searches many platforms and matches on skills, which surfaces qualified candidates beyond referral networks and single job boards. Selection still rests on job-related criteria and human decisions.

Is positive action legal in UK recruitment?

Yes, within limits. Under the Equality Act 2010, employers can take proportionate steps to encourage applications from under-represented groups. Preferring a candidate at selection is allowed only as a tie-breaker between equally qualified candidates.

Can employers prefer a candidate because of a protected characteristic?

Only in a narrow case. A general policy of preferring candidates with a protected characteristic is not permitted. Positive action at selection is limited to a proportionate tie-breaker between equally qualified candidates.

What are the protected characteristics under the Equality Act 2010?

The nine protected characteristics are age, disability, gender reassignment, marriage and civil partnership, pregnancy and maternity, race, religion or belief, sex, and sexual orientation.

Should sourcing tools use protected characteristics to rank candidates?

No. Sourcing tools should match on skills and job-related requirements. The ICO has asked AI recruitment providers not to filter or infer protected characteristics and to collect any monitoring data directly from candidates.