At Tailo, accessibility isn’t just a box-ticking exercise for us – it’s at the heart of everything we do. We’re always listening to feedback from our disabled and neurodivergent users to keep improving the way our assistive tech works. Our goal? To make our software as accessible, engaging, and empowering as possible.
Being part of the assistive tech world means we’re constantly keeping up with how AI is evolving – and what it could mean for disabled people. From AI-powered sign language interpreters to smart mobility aids and text-to-speech tools, the potential is huge.
But here’s the catch: AI can be biased. AI reflects the data it’s trained on, and when that data excludes or misrepresents disabled people, the tech ends up doing the same. That’s what disability bias is all about. So let’s talk about what it is, why it’s a problem, and what we can do to fix it.
So, what exactly is disability bias in AI?
Disability bias happens when AI systems are built using data that leaves disabled people out – or gets their needs completely wrong. This might be because the data doesn’t include enough (or any) disabled people, the design isn’t inclusive, or because developers didn’t bring people with lived experience into the process. Tools like facial recognition software or AI-driven hiring platforms for example often don’t work well for disabled people. Instead of breaking down barriers, they can end up creating new ones.
Even assistive tech isn’t immune – which is kind of ironic, since its purpose is to empower disabled people. For example, some speech-to-text apps don’t recognise the voices of people with conditions like ALS or cerebral palsy, and some tools meant for visually impaired users might struggle with identifying people with darker skin tones.
Why does it matter?
Because these problems go beyond just tech glitches. Bias in AI can lead to real-world discrimination in jobs, education, healthcare, and more.
It can reinforce the idea that “normal” (whatever that means) is the default, and disability is something out of the norm. As AI becomes more involved in decisions that affect our everyday lives, these biases risk becoming part of the system – harder to spot, harder to challenge, and harder to fix.
What can we do about it?
If we want to make AI truly inclusive, we need to rethink how we design and build tech, with disabled people involved in the development at every stage.
Here’s how:
- Use better data: AI needs training data that actually reflects the full range of human experiences, including physical, sensory, cognitive, and mental health disabilities – collected ethically and with consent.
- Design with accessibility in mind: Developers should build tech that works for everyone from the start, including things like screen readers, voice recognition and ability to change visual environments and formats.
- Co-create with disabled people: Include disabled voices at every stage, from the first idea through to testing and beyond. Feedback shouldn’t stop when the product launches.
- Test for bias: Run regular audits to check for bias and be open about the results. Make it easy for users to report issues and ask for improvements.
- Push for better policies: Governments and organisations need to create clear rules that make inclusive AI a must, not a maybe.
What we’re doing at Tailo
We’re committed to tackling disability bias in our software.
Here’s what we’re focusing on right now:
- Providing a multi-sensory reading experience, including human-sounding Text-To-Speech technology and customisable reading environments (adjustable colour overlays, fonts, and more!)
- Running user tests with disabled and neurodivergent people
- Designing with accessibility in mind from day one
- Regularly reviewing how we use AI – and making sure it’s responsible
We believe AI can be a game-changer for disabled people, but only if it’s built with them, not just for them. If you want to learn more about how Tailo is committed to inclusion and accessibility, find out more here.
Written by Mina Moriarty
Content Creator