Change Management in the Age of AI: Keeping Disabled Voices in the Conversation
There is a pattern in how change moves through our society. It arrives fast, reshapes everything it touches, and when the dust settles, we discover, again, that the people who needed the most preparation were given the least. People with disabilities know this pattern well. They have lived it through every major shift in technology, economy, and policy for generations. AI is no different. And if we're honest about what's happening right now, it may be more disruptive than anything that has come before. This is why we need to learn from our mistakes of the past, and make sure no one is left behind in an AI-enabled future.
The Scale
This is not one change. It is a convergence of many changes happening simultaneously. AI is altering the nature of work. It is reshaping how government services are delivered. It is changing how trust gets built (or broken) between institutions and the people who depend on them. Economic models are being stress-tested. Community infrastructure is adapting in real time. For most people, this is disorienting. For the approximately 61 million Americans living with a disability, it is a full-scale assault from every angle at once.
The stakes are not abstract. For someone navigating the job market with an intellectual or developmental disability, AI-powered hiring tools can quietly filter them out before a human ever reviews their application. For a person relying on government services, automated decision systems can create new bureaucratic walls without any of the accommodations built into the old ones. For direct support professionals (DSPs), the workforce that makes community living possible, rapid technological change can destabilize an already fragile employment pipeline. Change does not happen to systems. It happens to people. And when it hits this community, it tends to hit harder and leave fewer options for recovery.
Why We Get Left Behind
The honest answer is not malice. It is sequencing. When new technologies are designed, tested, and deployed, disability is almost never in the room. Accessibility gets treated as a compliance checkbox, not something you build toward from the beginning. This is a design failure masquerading as a resource limitation.
It also reflects a deeper structural issue: our community has historically been underrepresented in the data that trains AI systems, underrepresented in the policy conversations that govern them, and underrepresented in the workforce that builds them. The result is technology that does not see us and therefore cannot serve us.
How to Proactively Prepare for Change
The question worth asking is not how we respond after the wave hits. It is how we position our community to shape what the wave looks like before it arrives. That requires three things working together.
The first is education that is built for this community, not adapted for it. People with disabilities, their families, and the professionals who support them need AI literacy that is grounded in their actual lives, their work goals, their service systems, their rights, and their risks. Generic digital literacy programs are not enough. The training needs to be specific, practical, and interactive. It needs to make AI feel like a tool they can direct, not a force they have to survive.
The second is intentional inclusion in the design process. Organizations building AI-powered tools for employment, education, benefits administration, or health care need to go beyond accessibility compliance. They need disabled people in the design conversation, not as users to accommodate, but as co-creators whose lived experience improves the product for everyone. This is not a charity position. It is a quality argument. AI that works well for people with the highest support needs tends to work better across the board.
The third is advocacy infrastructure that can move at the speed of change. Policy in this space is moving fast, and the organizations that have historically championed disability rights are resource-constrained. Building the capacity to monitor, respond to, and shape AI policy with disability considered at every level is not optional anymore. It is a survival strategy.
The Workforce Dimension
One of the most underappreciated dimensions of this moment is what AI means for the direct support workforce. DSPs, job coaches, employment specialists, transition coordinators who translate system-level change into individual-level outcomes. When AI tools enter the organizations they work in, they often do so without adequate training, without clear guidance on how to use them responsibly, and without any input from the people doing the work. That is a change management failure that puts both the workforce and the people they serve at risk.
Getting this right means treating the support workforce as a primary audience for AI education, not an afterthought. It means investing in tools that reduce administrative burden without replacing the human judgment that the work requires. And it means creating space for direct support professionals to name what is working and what is not, because they will see the consequences of AI deployment long before the data does.
What We Owe Each Other
The disability community has always been at the forefront of demonstrating that designing for the margins makes things better at the center. Curb cuts. Closed captions. Voice interfaces. All of these originated as accommodations that became universal features. The AI era offers the same opportunity, if we are willing to insist on it.
That insistence has to be active. It has to show up in product decisions, in policy comments, in hiring practices, in training curricula, and in the everyday conversations happening inside organizations that are trying to figure out what this technology means for the people they serve. Keeping people with disabilities in the conversation is not a soft aspiration. It is a precondition for getting any of this right.
The wave is already moving. The question is whether our community is in front of it, or underneath it.