The Why in the AI
July 2025
Using AI to empower disabled jobseekers
Artificial intelligence (AI) holds immense potential to support disabled jobseekers in overcoming barriers to employment, yet its implementation remains disproportionately focused on macro-level policymaking rather than individual-level interventions. This paradoxical approach risks deepening the employment gap for disabled individuals, as systemic improvements fail to translate into tangible support for applicants during recruitment processes.
At the macro level, AI demonstrates extraordinary capabilities in refining policies. For instance, tools like “Humphrey” assist in consultations where vast amounts of feedback pour in when governments consider disability employment strategies. Humphrey can swiftly sort through millions of comments, distinguishing critical insights such as “reasonable adjustments” recommendations from suggestions on tax incentives, and identifying emerging trends to help policymakers act promptly. Similarly, AI contributes to high-level meetings where ministers and civil servants discuss strategies to reduce the persistent 30-point employment rate gap between disabled and non-disabled people. In these settings, AI delivers real-time digests, summarising ideas with cross-party support, highlighting recommended metrics, and pinpointing potential bottlenecks flagged by analysts. These applications illustrate AI’s ability to drive policy change effectively.
Unfortunately, this powerful technology often stalls when addressing the needs of disabled jobseekers directly. For candidates who self-identify as disabled, interactions with the recruitment system remain largely impersonal and underwhelming. For example, after submitting an application, disabled applicants often receive generic feedback—if any at all. There is currently no system in place to generate even a basic summary of why an applicant fell short during the selection stage or specific guidance on improving their interview performance. This lack of personalised feedback leaves individuals to navigate employment processes without meaningful support, undermining their ability to succeed and grow.
The UK’s Disability Confident scheme aims to go beyond mere non-discrimination by actively promoting the inclusion and success of disabled candidates. Among its promises are guarantees of interviews for any disabled applicant meeting minimum criteria, offering reasonable adjustments throughout recruitment, and providing constructive feedback to help candidates focus on areas for improvement. Yet, the lived reality often diverges from these commitments, with many disabled jobseekers left lacking the guidance and resources needed to thrive in competitive employment environments.
To truly level the playing field, AI must be integrated not only into policymaking but also into recruitment systems at the individual level. Imagine an AI-powered assistant that generates personalised feedback for each disabled applicant, outlining specific reasons for rejection and actionable steps to enhance future applications. Furthermore, AI could enable recruiters to deliver tailored recommendations for reasonable adjustments, ensuring all candidates can perform at their best. With such tools, AI would no longer be limited to shaping policies from a distance but could become a transformative force in empowering disabled jobseekers directly.
Leveraging AI to support disabled individuals during recruitment is essential to close the employment gap and fulfil promises like those of the Disability Confident scheme. By addressing gaps in individual-level interventions, AI has the potential to reshape the lived experiences of disabled jobseekers, ensuring that technology serves as a bridge to inclusion rather than a barrier.
We already have that capability in 2025, so why, on earth, can’t we use AI to help individual disabled applicants by summarising feedback, coaching on application weaknesses, or even suggesting reasonable adjustments in interview processes?
Because right now, all too ironically, we’re are heading to a situation where we are using AI to refine policies about disability employment, while leaving disabled applicants, say neurodivergent, non traditional, or “you’re not typical” candidates, to fend for themselves. That has the perverse effect of widening the very employment gap we claim to want to close.
