AI-Powered Communication Tools for Non-Verbal Clients: What Nonprofits Should Know

Understanding AAC and Who It Serves

Augmentative and alternative communication encompasses a wide range of strategies, tools, and technologies that support or replace natural speech. It spans a spectrum from low-tech approaches, such as picture boards and alphabet charts, to high-tech dedicated speech-generating devices costing thousands of dollars. The term "non-verbal" covers people who produce no speech, as well as people who produce speech that others find difficult to understand, people who communicate inconsistently due to fatigue or cognitive factors, and people who communicate in one context but not another.

The populations nonprofit organizations serve who may use or benefit from AAC are broad. Schools and educational nonprofits support children with autism, cerebral palsy, childhood apraxia of speech, and developmental language disorders. Disability services organizations work with adults across all of these conditions as well as acquired conditions such as ALS, multiple sclerosis, and traumatic brain injury. Hospice and palliative care providers serve people who lose the ability to speak in their final weeks or months. Healthcare nonprofits work with stroke survivors navigating aphasia, the language disorder that affects a person's ability to speak, read, write, and understand language.

Access to these tools has historically been deeply unequal. Research published in 2025 found that among minimally verbal children with autism, approximately 84% of white children had access to an AAC device, compared to roughly 32% of children from racial and ethnic minority families. More than half of minority families surveyed had no knowledge of AAC devices at all. This disparity is not simply a resource gap. It reflects differences in how families are informed about available options, how clinicians communicate across cultural and language barriers, and how funding systems favor those with more institutional knowledge. Nonprofits serving underrepresented communities have a particular role to play in closing this gap.

AI is not simply making existing AAC features faster. It is enabling capabilities that were not previously possible: personalized learning systems that adapt to individual users over time, voice reconstruction for people losing the ability to speak, recognition of non-standard and atypical speech, and conversational AI that can carry on exchanges rather than just deliver pre-programmed phrases. For nonprofits thinking about how to serve clients who need communication support, understanding these AI-specific capabilities is important.

Word prediction has been part of AAC for decades, but AI-powered prediction in modern systems is substantially more sophisticated. Traditional predictive systems relied on frequency and recency. Current systems use natural language processing and machine learning to understand context, conversation partner, location, time of day, and environmental cues to offer genuinely relevant predictions. Snap Core First from Tobii Dynavox, for example, predicts likely next words based on communication context, reducing the number of selections a user must make to construct a message. For users with motor impairments where every selection is physically effortful, this reduction in required steps has real functional impact.

However, clinicians and researchers have raised important questions about AI prediction in AAC. A 2025 paper in the ASHA Leader noted that predictive text may sometimes compromise user autonomy by steering communication in directions the user did not intend. When an AI system suggests complete phrases rather than individual words, the user may select a close-but-not-quite-right phrase because generating the alternative would require too many additional steps. This represents a genuine tension between efficiency and authentic expression that nonprofits supporting AAC users should understand.

Voice banking and voice cloning represent perhaps the most emotionally significant AI advance in the AAC space. When a person receives an ALS diagnosis, they often have months or years before they lose the ability to speak. Voice banking programs allow them to record speech samples that can be used to synthesize a personalized voice for their AAC device. Previously, this required hundreds or thousands of recorded sentences. AI voice cloning has reduced this requirement to as little as a few seconds of audio. ElevenLabs offers a free Impact Program specifically providing voice cloning technology to people with ALS/MND who are losing their ability to speak. Their technology can also handle what they call "slurred-to-clear" conversion, where a patient with advanced dysarthria speaks into a microphone and AI maps their impaired speech onto a high-fidelity voice clone in real-time. For hospice programs or ALS support organizations, connecting clients to these resources early in the disease progression is a meaningful contribution to their quality of life and autonomy.

Research consistently identifies that AAC device abandonment is not primarily a technology problem. It is an implementation problem. Devices are abandoned or underused not because they lack features but because the people who support AAC users, whether family members, direct care staff, paraeducators, or residential care workers, do not have the knowledge, time, or confidence to use them consistently. For nonprofit leaders, this means that any investment in AAC technology must be accompanied by investment in people.

The core practice that makes AAC effective is called aided language stimulation or modeling. It means that communication partners, the people interacting with AAC users, use the device themselves while speaking. They point to or activate symbols as they talk, demonstrating how the device is used and giving the AAC user a model to imitate. This practice is well-supported by research and is consistently cited as the single most impactful thing caregivers and staff can do to support AAC development. Yet in practice, staff who are new to AAC rarely model spontaneously, often because they don't know how, sometimes because they feel self-conscious, and frequently because their caseloads don't leave time to learn and maintain the skill.

A 2025 survey of approximately 500 speech-language pathologists found that 74% reported not feeling adequately prepared to provide AAC services. If trained SLPs feel underprepared, the teachers, paraeducators, and direct care workers who interact most frequently with AAC users are even further from confidence. Nonprofit organizations that take AAC seriously need to build ongoing training into their operational model rather than relying on one-time orientations. This connects to the broader challenge of building internal expertise, which is addressed in our article on building AI champions across your nonprofit.

Consistency across environments is another critical factor. AAC works best when the same device and vocabulary system travels with the user across all settings, home, school, clinic, community, and when all communication partners use it consistently. In organizational settings, this requires coordination across multiple staff members and programs. When a client uses Snap Core First at the day program but family members at home are using a different system, or when new staff don't receive AAC training, the communication environment becomes fragmented and progress stalls. Nonprofits that provide services across multiple settings have both the opportunity and the responsibility to coordinate this continuity.

For nonprofit leaders who serve populations that include people with communication disabilities, the path forward involves both immediate actions and longer-term strategic thinking. You don't need to be an AAC expert to make meaningful progress. What you do need is a willingness to invest in the right knowledge, relationships, and systems.

Start by assessing your current state. How many clients in your programs use or would benefit from AAC? Do the staff who work most directly with these clients have current knowledge of available tools and how to use them? Does your organization have a relationship with a qualified SLP who understands AAC? These baseline questions will reveal where the most urgent gaps are.

Build relationships with the relevant funding systems. If you serve school-age children, develop working relationships with school district SLPs and special education administrators. If you serve adults, identify your state's AT lending program coordinator and Vocational Rehabilitation district office contacts. If you serve people with progressive neurological conditions, connect with the relevant disease-specific organizations, such as the ALS Association, which can connect clients with voice preservation resources before they need them.

Invest in training with appropriate depth. One-time staff orientations are not sufficient for AAC. Organizations that achieve good outcomes build ongoing training into their operational model, including initial training for new staff, regular refresher training, and access to ongoing consultation when new challenges arise. The AAC Academy offers affordable online training that can build organizational capacity over time. Pairing this with AI-assisted onboarding approaches can help new staff build knowledge more efficiently.

Consider developing an equity lens for how you identify and address access disparities. If your client data shows that some populations, perhaps by race, language, income level, or diagnosis, are less likely to be referred for AAC evaluation or to receive devices, that disparity is worth examining. Proactive outreach, bilingual informational materials, and partnerships with community organizations that serve underrepresented populations can all contribute to more equitable access. This aligns with the broader organizational work described in our article on AI and nonprofit knowledge management.