Last updated:
Canada's federal regulatory framework (part of Bill C-27) aimed at ensuring high-impact AI systems are developed and used safely and without bias.
AIDA focuses on High-Impact AI and General-Purpose AI, requiring organizations to assess risks, implement mitigation plans, and provide public disclosures. It introduces the role of the AI and Data Commissioner with powers to audit organizations and issue significant penalties for non-compliance, aiming for a pro-innovation but human-centric balance.
Real world example:
A Canadian financial service provider inventories its credit-limit model as High-Impact under AIDA and establishes a specific bias-monitoring protocol and publishes a plain-language summary of how the model reaches decisions for Canadian consumers.
