At Alliance for Inclusion, we believe that technology should be a tool for progress, breaking down barriers, not building new ones.

 

Yet as Artificial Intelligence becomes increasingly central to recruitment and selection, it brings both opportunity and risk.

 

While AI offers efficiency and consistency for resourcing companies, it can also unintentionally exclude talented deaf, disabled and neurodivergent (DDN) candidates from fair consideration.

​

The Good and Bad of AI

​

From an efficiency standpoint, it’s clear that artificial intelligence has revolutionised recruitment. Automated tools can now screen thousands of CVs, analyse video interviews, scan candidates’ social media profiles, and even predict job performance. All at remarkable speed. The result is faster shortlisting, more data-driven insights, and, at least in theory, a reduction in human bias.

​

A good example comes from Unilever, where AI has been successfully integrated into the hiring process. The company has reported quicker recruitment times and significant cost savings, while allowing HR teams to focus on higher-value activities such as interviewing, onboarding, and employer branding. This demonstrates how AI can move beyond experimental “pilot” projects to deliver measurable impact in talent acquisition.

​

However, in practice, AI systems are trained on historical hiring data; and these datasets often reflect longstanding patterns of exclusion. As a result, candidates who are deaf, disabled, neurodivergent, or living with health conditions may be underrepresented. Algorithms trained on such biased data may wrongly categorise these individuals as high-risk rather than recognising the unique perspectives and value they bring.

​

Instead of addressing inequality, these technologies can end up reinforcing it, perpetuating existing barriers in recruitment. For instance, the AI-driven video interview platform HireVue faced major criticism for its facial analysis feature, which was found to disadvantage disabled applicants. Following public backlash and formal complaints, the company ultimately discontinued the use of facial analysis in its hiring process.

​

What might appear to be an objective, machine-made decision can, in reality, conceal deep-rooted human biases and historical patterns of exclusion.

​

Key Issues for Resourcing Companies

 

1. Algorithmic Bias

 

AI models are only as fair as the data they’re trained on. If that data favours candidates who fit a certain “typical” profile, then deaf, disabled and neurodivergent applicants, who may communicate or present differently, risk either being screened out before human review, or being lowered in ranking if the system provides recommendations.

 

2. Lack of Transparency

 

Many AI systems function as black boxes, providing little insight into how decisions are made. Recruiters may not be able to explain why a candidate was rejected, making it difficult to ensure fairness or compliance with equality legislation.

 

3. Legal and Ethical Compliance

 

Recruitment under the UK Equality Act 2010 must not disadvantage disabled candidates or those with any of the nine protected characteristics. AI tools that assess voice tone, eye contact, or typing speed risk breaching these requirements if they penalise such differences.

 

4. Reputational and Brand Impact

 

A lack of transparency or accessibility in hiring technologies can harm an organisation’s reputation. Inclusive recruitment is not just a compliance issue; it’s a brand value. Candidates and clients alike expect fairness, accountability, and diversity.

 

5. Process

 

AI may not identify if a candidate is neurodivergent or not, or if they are identified, ensure processes are adapted to be more accessible, such as avoiding the use of psychometrics or other behavioural testing that can lead to discrimination because it was only designed for neurotypical responses.

 

 

What is it like for DDN Applicants?

 

For many DDN people, AI-driven recruitment introduces new obstacles. Online assessments may not be compatible with assistive technologies such as screen readers. Timed cognitive or motor tasks can disadvantage candidates who require additional time or adapted formats.

 

Video interviews pose further challenges. Algorithms assessing facial expressions, tone, or body language might misinterpret communication differences linked to disability or neurodiversity. Often, candidates are rejected automatically, without any opportunity for human review or reasonable adjustments.

 

This automated rejection process can be disheartening and isolating, leaving talented candidates excluded from roles they are fully capable of performing.

 

 

Building Inclusive AI in Recruitment

 

Resourcing companies have a crucial role to play in shaping fairer recruitment systems. Practical steps include:

​

• Bias Auditing: Regularly test and monitor AI tools for discriminatory outcomes.

• Accessible Design: Ensure platforms work with assistive technologies and offer alternative formats or assessments.

• Human Oversight: Keep humans in the decision-making loop to review automated outcomes.

• Transparency: Explain how AI is used and offer meaningful feedback to candidates.

• Co-design with Disabled People: Engage disabled professionals and advocacy groups in system testing and development.

​

Striking the Balance

 

AI has the power to make recruitment more efficient and insightful, but efficiency must never come at the cost of equality.

 

At Alliance for Inclusion, we champion a future where technology and inclusion go hand in hand.

 

Recruitment should reflect the diversity of the workforce it seeks to build, ensuring every candidate, DDN or not, has a fair chance to succeed.

 

Building inclusive AI is not just about compliance; it’s about fairness, humanity, and opportunity. When designed responsibly, AI can become a true enabler, helping to create workplaces that value difference and give everyone the chance to thrive.