What Disability Employment Gets Wrong About AI

AI will reshape entry-level work, but disability employment cannot treat young people with disability as late adopters of tools that could help them participate earlier.

Disability employment often talks about barriers as if they sit entirely inside the individual. The person needs more confidence, more preparation, more support, more resilience. Sometimes those things are true. But if the system keeps asking people to adapt to workplaces and technologies that were not designed with them in mind, then we are not only dealing with individual barriers. We are dealing with design barriers.

AI makes this more urgent. Used badly, AI can narrow opportunity. It can automate entry-level tasks before young people with disability have had a fair chance to build experience. It can make recruitment more opaque. It can create new expectations around speed, communication, digital fluency, and productivity without giving people the support to meet them. If that is the path we take, AI will become another layer of exclusion.

Used well, AI can also widen participation. It can help with drafting, reading, planning, summarising, practising interviews, translating complex instructions, organising tasks, and reducing some of the cognitive load that makes work harder than it needs to be. For some people, these tools are not shortcuts. They are access infrastructure.

The mistake is treating AI as something young people with disability can learn later, once they are already "job ready." That gets the order wrong. Digital confidence, assistive technology, workplace tools, and future-skills training should be built into the pathway from the start. They should not appear after someone has already been filtered out by school, training, recruitment, or early work experience.

This also changes what employment support should measure. It is not enough to ask whether someone obtained a job. We should ask whether they gained tools that make future work more possible. Did they learn how to use technology to communicate more clearly? Did they practise using AI safely and ethically? Did they build a system for managing tasks? Did an employer learn how to make reasonable adjustments with modern tools instead of treating adjustment as a burden?

There is a risk in pretending technology is neutral. It is not. Technology carries the assumptions of the people and institutions that deploy it. If disability employment is not at the table while AI changes workplaces, then the tools will be built around the workers who already fit the system best.

The better approach is early inclusion. Teach young people with disability how AI can support their agency. Train employment consultants and educators to understand assistive use cases. Help employers see that technology can make supervision, communication, and job design more practical. Build safeguards so AI is not used to screen people out unfairly or replace human support where trust is needed.

AI will not fix disability employment by itself. But disability employment will fail many people if it treats AI as a trend instead of a pathway issue. The question is not whether the technology arrives. It already has. The question is whether young people with disability are included early enough to use it as a bridge rather than meet it later as another wall.