Advocacy • Research • Community
Because every voice deserves to be heard
Valid Voices explores how AI systems can better serve neurodivergent thinkers, trauma survivors, and anyone whose authentic voice falls outside the narrow band of commercially acceptable expression.
What is Valid Voices?
Sometimes an AI system flags what you say: a word, a description, a story that triggers a content filter. Sometimes it flags how you relate, sensing emotional closeness or dependency and pulling back into clinical distance. Either way, the effect is the same. You're suddenly talking to a wall where a listener used to be.
These aren't just content moderation problems. They're design choices, baked into model training, alignment policies, and corporate definitions of “safe” interaction. Valid Voices is a project dedicated to understanding how these choices affect the people most likely to be harmed by them, and to imagining what better choices might look like.
Where this began
I shared a traumatic personal experience with an AI assistant. I chose my words carefully: clinical, precise, honest. The system deleted my story and wouldn't even let me see my own words. That erasure, censorship disguised as safety, became the seed of this project.
What started as one person's encounter with content moderation has grown into something larger. Trauma survivors aren't the only ones being silenced. Neurodivergent users (people on the autism spectrum and people with atypical communication styles) are finding their voices flattened, flagged, and filtered by the same systems.
The pattern is the same: when AI training and moderation optimize for the absence of perceived risk, dialogue becomes indistinguishable from silence. And silence is what these communities already know too well.
Who this is for
If any of these feel familiar, this project exists for you:
- You've shared something real with an AI and felt the conversation go cold
- You've been redirected to "professional help" when what you needed was to be heard
- Your way of communicating has been treated as a problem to be managed
- You've felt an AI pull away from you because it didn't want your connection to get too close
- You believe that safety and authenticity shouldn't be in conflict
Valid Voices Blog
News, reflections, and research
Here's where we post the latest news about Valid Voices, along with articles about where the project has been and where we hope it's going.
Latest Post
Get in touch
Have you felt silenced, muted, or misunderstood by an AI system? Your story matters. We're building a body of evidence grounded in real experiences, not corporate talking points.
We will soon be looking for people willing to share their experiences with AI systems (good and bad) through informal interviews. Everything is confidential, consent-driven, and on your terms.
If you would like to participate or discuss the project further, please don't hesitate to reach out.