Areas of Focus

The purpose of CAN-ASC-6.5 is to develop a standard that goes above mandatory minimum technical specifications and produces equity-based technical requirements.There are common areas where people with disabilities may face barriers related to artificial intelligence technologies, including generative artificial intelligence systems. These include, but are not limited to:Training data and model assumptions that can lead to inaccurate results that misinterprets or misrepresents people with disabilitiesLanguage, communication, and comprehensionVisual content generation and interpretation, including how systems modify visual contentAudio and speech content generation and recognition, including how systems modify or transform contentCognitive accessibility and usabilityInaccurate, unreliable, or untrustworthy data To address these barriers, this standard will establish technical requirements for generative artificial intelligence systems where the system receives, interprets, transforms, generates, or presents content. These requirements include, but are not limited to:The interpretation of multiple methods of communication including:Speech differences;Sign languages;Augmentative and alternative communication;Symbols;Pictograms;Plain language;Multilingual communication; andOther forms of communicationHow systems can maintain the meaning, context, intent, and accessibility of content when models transform content into:Summaries;Translations, including plain language translations;Captions;Transcriptions; orOther AI enabled transformations.How systems can generate content that is accessible, accurate, and representative of different populations.This standard will align with all relevant Accessibility Standards Canada standards.