When technology gives you your voice back
Designing dignity and inclusion through AI: Google’s Project Euphonia
When we design for accessibility, we’re not helping a few but for everyone
I’ve always believed that technology should feel human, but studying Google’s Project Euphonia made me realize what that really means.
For many people living with ALS, cerebral palsy, or even throat cancer, losing their voice doesn’t just mean losing a way to speak; rather, it means losing a part of who they are. Voice carries identity, tone, humor, and emotion. When a robotic, generic sound replaces that, communication becomes functional but not personal. One of the best examples to provide is Dr. Stephan Hawking, with his robotic-sounding voice, though iconic, also became a symbol of limitation.
Project Euphonia changes the way I think about accessibility. It’s an AI-powered initiative that helps people with speech impairments communicate more naturally and, most importantly, in a voice that still sounds like them.
Designing dignity, not just functionality
What stuck me most is how this project reframes accessibility. Instead of seeing disability as a limitation, it recognizes that the real barriers are often the design itself. The issue isn’t that someone can’t speak, but that it’s that the tools they’re given don’t reflect who they are.
When people can record samples of their natural speech and later use that model as their voice, it’s not just practical but also emotional. It preserves identity. It gives dignity back to communication. As designers, we often talk about empathy, but this project shows what empathy looks like in action.
When technology learns to listen
Another layer that stayed with me was how Project Euphonia trains AI to understand atypical speech. Most voice assistants fail to recognize slurred or non-standard patterns, which can make users feel unseen and unheard. Coming from personal experience here, whenever my father, who is personally very fascinated with voice assistants, tries to use any, because of his dialect, most of them are not able to respond to him, or sometimes even detect his voice. I think this is a feature that’s not just transforming for accessibility but in a way also evolving abled-focused technology.
Project Euphonia adapts; it listens better, learns better, and respects that person on the other side. It reminded me that inclusion isn’t about adding more features but about removing friction where people feel excluded.
My eureka moment
While watching real users share their stories, I was honestly emotional. A person with ALS uses their preserved voice to talk to family. Like a grandparent recording family stories in their real tone… It’s hard not to feel the power of that.
Hearing those voices made me think about how design goes beyond usability. It’s about memory, presence, and connection. That’s what good design is doing here: it's restoring humanity.
Although…
I think there’s a deeper, more complex side to all this. Voice cloning technology can easily be misused. It can sound “almost human” but still lack warmth, or even worse, be used for harm.
It made me realize that designing for accessibility also means designing for responsibility. Transparency, consent, and user control aren’t optional here; in fact, they’re moral necessities. The technology has to protect people as much as it empowers them.
As a Designer?
I’ve come to see accessibility differently. It’s not just a design checklist or a side feature; it’s the foundation of meaningful design.
A voice isn’t just sound but a story.
When we build technology that helps someone reclaim that story, we’re not just solving a problem but giving them back a piece of themselves. Project Euphonia reminded me that the best kind of design doesn’t just make life easier, it makes it more human.
References:
Google Research — Project Euphonia: https://sites.research.google/euphonia/about/
ALS Association: https://www.als.org/understanding-als/what-is-als
