AIDA is coming: what Ontario AI shops should draft now
Canada's Artificial Intelligence and Data Act will make automated-decision liability real. Here's the three policies and one endorsement to put in place before it lands.
What AIDA actually does
AIDA is Canada's proposed Artificial Intelligence and Data Act, which will regulate 'high-impact' AI systems and create liability for harms caused by automated decisions. It follows the EU AI Act's risk-tiering logic but adds Canadian administrative-penalty teeth.
For Ontario AI shops, the operative effect is that shipping a high-impact system without the required assessments, documentation, and output logs becomes a compliance failure with a financial exposure - and, in serious cases, a director-level exposure.
The 'high-impact' line
The line that matters is 'high-impact.' Systems that affect employment decisions, credit, health, or essential services are squarely high-impact. A general-purpose LLM wrapped into a customer-support bot may or may not be - but the moment it influences a hiring screen or a benefits decision, it crosses.
The practical move: classify every shipped AI feature against the high-impact criteria before AIDA's enforcement date, and document the classification. The documentation is what an AIDA examination will ask for first.
Reading is the start. A short conversation turns it into a plan.
No sales pressure. We'll look at your stage, your contracts, and your AI exposure, and tell you - plainly - where your current coverage won't hold up.
Risk notes, once a month.
One short note on Ontario tech risk - breaches, regulation, and what your policy should actually cover.
Three policies to write first
Write three before the law lands: an AI development standard, a model-output review process, and an incident-response plan specific to AI. The development standard controls training-data provenance; the review process gates what ships; the incident plan covers both a regulator examination and a customer harmed by model output.
On the insurance side, confirm your AI E&O endorsement covers administrative penalties from an AIDA examination and extends the definition of professional services to include AI development. Without it, the compliance fine and the defence of the examination may both sit on your balance sheet.
Keep reading
Ransomware is changing in 2026 - and Ontario underwriters know it
Encryption is the old playbook. Extortion - the threat of leaking data, not locking it - is the new one. Here's what that means for your cyber limit.
PHIPA breach windows: what actually counts as 'prompt'
Ontario's IPC has tightened what 'prompt notification' means under PHIPA. The window is shorter than most healthtech founders think - and the fines are personal.
When your AI hallucinates, who pays?
The Air Canada chatbot case is now precedent Ontario tribunals cite. If you ship an LLM feature, your Tech E&O needs an AI endorsement - not a hope.
Keep exploring
General information only. Not insurance advice.