Google to roll out AI age estimation in Canada as Ottawa weighs Bill C-34
Google will begin gradually rolling out its artificial intelligence age-assurance technology in Canada in the coming days while it monitors progress. The system is designed to detect underage users and apply additional protections to their accounts. The Globe and Mail
The rollout will be gradual. Google's age-estimation model uses machine learning to interpret signals already associated with a user's account. Those signals include types of information searched for and categories of YouTube videos watched. Google said it will not collect new information beyond what is already associated with the account for age assurance.
Google published its Canadian announcement on its official blog under the title "Ensuring safer online experiences for Canadian kids and teens". Google Canada blog The company says when its tools estimate a user might be younger than 18, it will notify them by email and as they use Google products. If the technology identifies a user as underage, the account will be changed to include additional protections such as safeguards on content recommendations. An adult mistakenly identified as under 18 can prove age using a government ID or a selfie.
Google Canada's director of government affairs and public policy is Jeanette Patell. Her portfolio covers the file in Ottawa as the company navigates federal interest in online safety and provincial concerns about youth privacy, education and child welfare.
Meta is moving on a parallel track. It has already started using AI in Canada to identify underage users and place them in teen accounts. Meta said in September it would expand testing of its AI age-identification technology to countries outside the United States, including Canada. The Globe and Mail
Its methods rely on account behaviour already on file. Meta uses AI to examine context clues such as birthday celebrations to identify accounts that may be underage. In May, Meta said it would add visual analysis involving AI scanning photos and videos for visual clues to a user's age. Persona, an identity vendor, is separately testing different scenarios teenagers may use to appear older than they are. Reuters
The U.S. legal picture does not map directly onto Canada. A settlement allows Meta to temporarily use some data generated by children it identifies in order to train an AI model to detect under-13s. Reuters Meta has confirmed that child protection measures promised as part of a recent U.S. legal settlement will not be coming to Canada.
In Ottawa, statute is lagging company policy. Canada's Liberal government introduced digital safety Bill C-34, which has not yet become law. The bill would require social media platforms to at least temporarily block access for children under 16. Parliament has not passed it.
The broader context here is familiar to federal watchers. Voluntary, Canada-wide technical systems are now operating ahead of a federal standard. Ottawa will have to decide whether it favours industry-led estimation with appeal routes, or a legislated access rule with a fixed age threshold. Provinces will weigh the same tools through their own lenses of schools, child protection and privacy enforcement.
Looking at what this means for Parliament, the questions are practical rather than abstract. Lawmakers will need to test accuracy and recourse. How often are adults misclassified, and how burdensome is ID or selfie verification. They will need to test data minimisation claims against what the models actually ingest. And they will need to determine whether content safeguards for under-18 accounts meet the stricter test in Bill C-34 of at least temporary blocking for under-16s. Those answers will shape committee study, regulation and any federal-provincial negotiation over enforcement.
