AI-empowered Project Report
Executive Summary
The AI-empowered research project examines how artificial intelligence (AI) and automation are impacting the employment opportunities and working experiences of disabled people in the UK. The research, undertaken between October 2025 and May 2026, combines a review of existing evidence with surveys of 1,029 employers and 1,402 disabled people, 115 in-depth interviews, and five organisational case studies involving 92 disabled people.
Key findings
AI presents both significant opportunities and risks for disabled people. AI has the potential to empower through increased flexibility, personalising ways of working, improving accessibility, supporting and enhancing assistive technologies, helping to address skills gaps and enabling disabled people to perform tasks in ways better suited to individual needs.
AI benefits are not automatic and require intentional application of inclusive design principles. AI adoption should focus on supporting people to do their jobs better, not replacing them. Without appropriate training, human support and effective governance, AI risks making existing inequalities worse, creating new barriers and increasing discrimination and exclusion.
AI adoption cannot resolve the underlying inequalities faced by disabled people in the labour market. Existing discrimination, inaccessible recruitment, inconsistent provision of reasonable adjustments, limited organisational understanding of disability, and difficulties accessing employment support shape employment outcomes independently of AI.
The research therefore argues that AI adoption must be accompanied by improvements in disability inclusion, workplace practice and employment policy. Work provides purpose, meaning, well-being and opportunities for contribution as well as economic value. AI adoption should therefore be judged against its wider social consequences, not simply productivity gains.
Eight principles for disability-inclusive responsible AI
The report concludes with eight evidence-led principles for Government and organisations. These call for action to (1) address existing employment inequalities; (2) recognise the diversity and value of disabled people; (3) develop digital and AI skills; (4) prioritise AI strategies that focus on AI enhancing human-led tasks over automation and replacement; (5) adopt accessibility-first approaches; (6) strengthen organisational leadership, accountability, and governance; (7) use personalisation benefits of AI to improve productivity and performance, and (8) ensure that social value guides AI adoption.
The report outlines 21 recommended actions for UK Government, including strengthening employment support and Access to Work, reviewing Disability Confident, promoting reasonable adjustments passports, involving disabled people in AI policymaking, researching the impact of AI recruitment systems, and updating workplace support frameworks to reflect developments in AI.
Finally, the report introduces the AI-empowered employer toolkit, designed to help organisations assess current policies, practices and capabilities, and identify context-specific actions to enhance disability inclusion while implementing AI responsibly.
Overall, the central message is that AI should be used as an empowering tool, not as a technological solution to disability. Its benefits will depend on whether organisations and policymakers address existing inequalities, involve disabled people meaningfully, prioritise accessibility and maintain appropriate human support. The research provides a framework for achieving this while recognising both the transformative potential of AI and the risks it presents.