For OCAD University Professor Dr. Jutta Treviranus, the release of the first global standard on accessible and equitable artificial intelligence (AI) systems is a significant achievement in helping to ensure that AI systems are accessible, inclusive and fair from the start.
“This new standard aims to protect people from exclusion and ensures that AI works for everyone,” explains Dr. Treviranus, director of the OCAD U-affiliated Inclusive Design Research Centre, who chaired the Technical Committee that developed the standard.
Released by Accessibility Standards Canada on Dec. 3, 2025, the new standard is free and open and was developed by a committee of people with lived experiences of barriers and opportunities created by AI. It also reflects input from international experts and international disability organizations as well as the public.
“Clear accessibility guidelines, as reflected in this new standard, can help ensure that barriers in AI are recognized and addressed before they impact users, especially for people with disabilities,” says Dr. Treviranus.
The new standard, CAN-ASC-6.2 Accessible and Equitable Artificial Intelligence Systems, can help organizations and developers design AI systems that are accessible to people with disabilities; ensure fairness and prevent exclusion; establish processes so that accessibility and equity remain central throughout the AI lifecycle; and educate others on creating equitable and inclusive AI.
“AI is shaping how Canadians access services, information, and opportunities. We want to ensure AI works for everyone, including people with disabilities. It’s essential if we want to build a fair and inclusive future. This new standard fills a critical gap by providing clear and practical guidance to help organizations design AI systems that are accessible and equitable from the start. With this publication, Canada continues to show leadership in accessible technology on the global stage,” says Dino Zuppa, CEO of Accessibility Standards Canada.
The standard is supported by capacity building resources and an inclusive gathering to enable informed critique by individuals and communities that will feel the impact of the standard.
Dr. Treviranus notes that this is the first phase of the standard, which outlines a framework and addresses pervasively deployed automated decision systems that predate the large language models and GenAI that have gained public awareness, such as ChatGPT. The new standard addresses the gaps in other AI guidance with respect to the risks and opportunities specific to disability.
Going forward, she says the Technical Committee plans to develop a standard that includes AI Authoring with the Web Accessibility of the W3C, address cybersecurity; and address the risks and opportunities of language models and generative AI for people who experience communication barriers.