Canada has published a new national standard aimed at ensuring artificial intelligence systems are accessible and equitable for people with disabilities, setting out one of the most comprehensive frameworks to date for inclusive AI design, deployment, and governance. The standard, CAN-ASC-6.2:2025 – Accessible and Equitable Artificial Intelligence Systems, was released in December by Accessibility Standards Canada and is intended to guide both public- and private-sector organizations that develop or use AI technologies.
The voluntary standard establishes requirements to ensure that people with disabilities can participate fully in the AI lifecycle—as creators, deployers, overseers, and end users—while also protecting them from disproportionate harms. It applies across the full lifecycle of AI systems, from planning and data collection to procurement, deployment, monitoring, and eventual termination.
At its core, the standard emphasizes four principles: people with disabilities should experience equitable benefits from AI systems; they should not face inequitable harms; their rights and freedoms must be preserved; and they must retain agency and be treated with dignity. To achieve this, organizations are required to embed accessibility and equity considerations from the outset, rather than retrofitting systems after deployment.
The document calls for active and compensated involvement of people with disabilities in AI design, testing, procurement, and oversight. It also requires organizations to assess and mitigate risks such as bias, statistical discrimination, cumulative harms, loss of privacy, and exclusion from essential services. Particular attention is given to the risks faced by people with disabilities as statistical outliers in data-driven systems, even when datasets appear representative.
The standard introduces detailed governance expectations, including accessible transparency and explainability documentation, consent and opt-out mechanisms, public registries of harms and contested decisions, and clear accountability for AI-assisted outcomes. Organizations must also provide equivalent non-AI or human-in-the-loop alternatives where AI systems are used to make or inform decisions that affect people with disabilities.
In addition, the framework supports broader systemic change by requiring accessible AI education, training, and literacy initiatives, developed with the participation of people with disabilities. The standard aligns with the Accessible Canada Act, the Canadian Human Rights Act, the United Nations Convention on the Rights of Persons with Disabilities, and international AI governance efforts.
Accessibility Standards Canada said the standard is designed to push innovation beyond minimum compliance, encouraging organizations to adopt equity-based approaches that improve outcomes for people with disabilities while delivering wider social and economic benefits.
Need Help?
If you have questions or concerns about how to navigate the global AI regulatory landscape, don’t hesitate to reach out to BABL AI. Their Audit Experts can offer valuable insight, and ensure you’re informed and compliant.
