Yes — Canada’s anti-spam law is written to be technology-neutral. The risk AI adds is scale outrunning the parts of the law that were fine at a smaller volume.
Short answer
Yes. CASL regulates the commercial electronic message — consent, sender identification and an unsubscribe mechanism — regardless of whether a person or an AI tool drafted or sent it. Nothing in the Act singles out AI-generated content for extra restriction. What AI actually changes is the risk profile: it makes it easy to send further, faster and more personalized than the underlying consent record or your compliance process was built to handle.
Section 6(1)–(2) of CASL sets out the same three-part test for every commercial electronic message: valid consent (express or implied), accurate sender and contact identification, and a working unsubscribe mechanism. The statute does not mention drafting tools, mail-merge, or AI anywhere — it was written to catch the message, not the method. Treadstone Law’s CASL email-marketing overview covers the same three-part test for a general audience.
Three places an AI-run email programme drifts off CASL faster than a manually run one usually does. First, contact information has to stay “valid for a minimum of 60 days” after a message is sent (CASL s.6(3)) — an AI tool auto-populating a sender address from a template can quietly break this if the underlying inbox changes. Second, unsubscribe handling has to keep pace with volume: the mechanism must let a recipient opt out using “the same electronic means by which the message was sent” and be honoured within 10 business days (CASL s.11(1), (3)) — a send volume AI makes possible but a suppression list a human still has to maintain correctly. Third, AI-assisted personalization at scale makes it easier to quietly send beyond the list that actually has valid consent for that purpose. Treadstone Law’s unsubscribe-mechanism guide sets out what a compliant mechanism actually needs to do.
If an AI tool is used to write the marketing copy itself, a separate law applies to what the email claims. Under the Competition Act, s.74.01(1)(b), making a representation about a product’s “performance, efficacy or length of life” that is “not based on an adequate and proper test” is reviewable conduct — and the burden of having run that test sits with the business making the claim, not with a regulator to disprove it. An AI copywriting tool doesn’t carry that burden for you; it can produce a confident-sounding performance claim with no test behind it at all, and the business that sent it is the one on the hook.
CASL s.20(4) caps an administrative monetary penalty at $10,000,000 for an organization, and a CASL complaint moves through the CRTC’s enforcement process the same way whether the message that triggered it was written by a person or a model.
For the consent question specifically, see whether existing consent covers AI-sent messages.
See how AI fits into a compliant growth and marketing engine.
