Skip to main content
All stories

A Disability Tech Founder Describes a Child's New Way to Say No

Source: Trevor Nichols, BetaKit, 19 November 2026

A child in Australia who previously had only a signal for yes gained the ability to say no while testing technology from Possibility Neurotechnologies, according to Pete Horsley, founder of Australian disability technology accelerator Remarkable. Speaking at the Canadian Disability Tech Summit in Toronto, Horsley said the child then used that new option frequently.

Trevor Nichols’ reporting for BetaKit places that account alongside a less reassuring assessment of AI accessibility: systems built around statistical patterns can serve people with disabilities poorly when their needs fall outside those patterns. The report, published on 19 November, covers a panel at Access to Success’ inaugural summit.

A communication gain, with limited evidence

Possibility, a Calgary-based brain-computer interface company, aims to let people control devices using their thoughts. Horsley described an Australian test in which a child acquired another way to communicate. His account gives the technology a specific purpose: enabling a response the child previously could not signal through the means he described.

BetaKit does not provide the child’s age, the test protocol, the device’s accuracy or evidence about how long the benefit lasted. The reported gain therefore remains an individual account. It cannot establish how reliably the technology would work for other children or in everyday use.

Voice restoration offers another application. Max Lemmens, ElevenLabs’ general manager for Canada, described using AI and old recordings to help people living with permanent voice loss regain access to their own voice. “For many people, their voice is their identity,” he said. BetaKit explicitly attributes the company’s reported reach to Lemmens; it does not present an independent assessment of those outcomes.

The panel also discussed access during emergencies. Tom Butcher, a senior advisor and ambassador at disability research organisation Zero Project, cited the World Institute on Disability’s use of AI to compile information about wheelchairs, interpreters and accessible shelters during crises. This is a different task from restoring communication: helping people locate resources they need.

When the average user is the wrong starting point

Jutta Treviranus, a professor at OCAD University and director of the Inclusive Design Research Centre, questioned the assumptions beneath such tools. She argued that AI’s preference for common patterns can produce wrong determinations about people with disabilities. Limited data about disabled users compounds the problem, according to her account.

Her proposed response involves re-engineering algorithms and enabling AI to run on a person’s device using individual data. Lemmens suggested another approach: synthetic training data based on disability-focused datasets. These were proposals discussed by the panellists, not solutions demonstrated by evidence in the report. Nichols’ account provides no comparison showing which approach better serves particular access needs.

Horsley also described more founders with disabilities starting businesses using AI tools, and argued for building accessibility into new systems rather than adding it later. He said retrofitting for accessibility is typically costlier. His argument also gave people with disabilities a role in creating those systems, rather than simply being asked to use them. As he put it at the summit: “Disability has always been a driver of innovation.”

Nichols’ full BetaKit report contains the speakers’ accounts and their differing proposals for making AI more accessible.

Source

This story is based on Trevor Nichols, BetaKit, 19 November 2026. Read the full original for the complete detail.

Read the original source →