Google to begin identifying underage users in Canada using AI

The Globe and Mail · collected 2026-09-14 · by Anja Karadeglija
Read the original at The Globe and Mail ↗

Summary

Google plans to introduce AI technology in Canada to identify underage users by analyzing existing user data like search history and video views. If identified as under 18, their accounts will receive additional protections; adults incorrectly flagged can appeal with government ID or selfies. Jeanette Patell from Google says the company will roll out this system gradually while monitoring its effectiveness. The move aligns with growing governmental concerns over online safety for minors and follows similar initiatives by Meta in Canada.
Written by the local model on 2026-09-14, using this article's own text rather than the other coverage of the same event (that is the story summary below).

Signals How these are calculated →

Claims extracted
23
claim-shaped sentences
Uncertain
9%
2 of 23 hedged
Leaning
not political
takes no side on a contested political question
Correction & hedging signals
40.0
corrections and hedging in what we collected; not a measure of accuracy
Outlets on this story
1
Technology
Narrative spread
1
articles carrying this framing
Analyzed 2026-09-14 · how these are computed

AI analysis (generated at analysis time, not now)

Story summary

Google plans to implement an AI-based age estimation model in Canada to identify underage users. The technology uses machine learning to analyze existing data from a user's Google account, such as search history and YouTube video categories. If the system identifies someone under 18, it will apply additional protections like restricting certain content recommendations. Adults incorrectly flagged as minors can verify their age with documents or selfies. Jeanette Patell, Google Canada’s director of government affairs and public policy, announced that the technology will be rolled out gradually in Canada over the coming days, with no new data collection beyond what is already associated with user accounts.

Written for “Google AI Underage User Identificatio…” on 2026-09-14, grounded in this article and the 0 other(s) covering the same event.
Why this leaning score
This article does not take a side on a contested political question, so it has no leaning score. That is an answer rather than a gap: a match report or a rescue can be warmly or critically written without being left or right, and scoring it anyway is how approval of a subject gets recorded as a political position.
No political leaning scored for article 8680 · logged 2026-09-14

Story

📰 Google AI Underage User Identificatio…
Technology · 1 article(s) covering the same event. This is the one the site leads with.

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Compared with similar articles

Nothing to compare against. No article is close enough to this one for the pipeline to have linked or judged the pair.

Publisher

The Globe and Mail · 174 article(s) · 14 correction(s) detected
Running correction rate · 14 correction(s)
2026-09-14
Morning Update: Broadening our economic horizons
2026-09-11
Morning Update: Gander remembers, 25 years later
2026-09-10
Morning Update: A long wait for trade talks
2026-09-09
Morning Update: Adversity ahead
2026-09-08
Morning Update: Say hello to Canadian countertariffs
2026-09-07
Morning Update: Tumbler Ridge survivors file new lawsuits
2026-09-06
Morning Update: Made-in-Canada trains
2026-08-31
Ontario plan to stop mandatory inquests for inmate deaths a mistake, advocates say
2026-08-31
Morning Update: Mapping new points of resistance
2026-08-28
Morning Update: Lake Ontario forever
2026-08-27
Morning Update: Rethinking resources as a trade weapon
2026-08-26
Morning Update: Ottawa bites back
2026-08-25
Morning Update: A new trade deal remains in the cards
2026-08-24
Morning Update: ‘You are at war when you get attacked’

Who wrote this

Anja Karadeglija
1 article(s) here · 1 carrying a prediction
🔮 The company says that if the technology identifies someone as underage, their Google account will be changed to include additional protections, such as safeguards on content recommendations.
The only article under this byline in the corpus.

Topics

Canada Google Google GOOGL-Q Meta YouTube

Subjects

Canada GPE · 6× Google ORG · 2× Meta ORG · 2× Google Canada ORG · 1× Google GOOGL-Q ORG · 1× Jeanette Patell PERSON · 1× Meta META-Q ORG · 1× Patell ORG · 1× YouTube ORG · 1× the United States. GPE · 1×

Narrative

“Our age estimation model uses machine learning to interpret a variety of signals already associated with a user’s account, such as the types of information a user has searched for or the categories of videos they’ve watched on YouTube,” the company said in a blog post Monday.
framing: assertive · carried by 1 article(s) · first seen 2026-09-14
🔮 The company says that if the technology identifies someone as underage, their Google account will be changed to include additional protections, such as safeguards on content recommendations.
2026-09-14 · The Globe and Mail
Google to begin identifying underage users in Canada using AI · assertive framing

Claims (23 extracted, 2 hedged)

Google GOOGL-Q is bringing its age assurance technology – which uses artificial intelligence to detect underage users – to Canada. asserted
which → bring → Canada
“Our age estimation model uses machine learning to interpret a variety of signals already associated with a user’s account, such as the types of information a user has searched for or the categories of videos they’ve watched on YouTube,” the company said in a blog post Monday. asserted
company → use → post
The company says that if the technology identifies someone as underage, their Google account will be changed to include additional protections, such as safeguards on content recommendations. asserted
account → say → recommendations
If it mistakenly identifies an adult as being under 18, the company says that person will be able to prove their age using things like a government ID or a selfie. asserted
person → identify → ID
“In the coming days, we’ll begin to gradually roll out this technology in Canada, carefully monitoring progress as we expand its availability,” Jeanette Patell, director of government affairs and public policy at Google Canada, said in the post. asserted
Patell → come → post
She said the company won’t be collecting any new information that isn’t already associated with the user’s account. asserted
that → say → account
“Age assurance technology is becoming more widely available, and we’re encouraged to see its increasing adoption by other companies in recent months,” Patell wrote. asserted
Patell → become → months
“While some would like a system that imposes a universal arbiter of age, we believe that liability and responsibility rest with every service owner – the developer, the publisher, the app creator – because they know their users and services best.” asserted
they → like → users
The increasing use of age assurance technology comes as governments around the world, including in Canada, move to implement age restrictions for social media. Meta META-Q has already started using AI to identify underage users and place them in teen accounts in Canada. asserted
Q → increase → Canada
The company said in September it would expand testing of the technology to countries outside the United States., including Canada. asserted
it → say → Canada
Meta uses AI to look at context clues, such as birthday celebrations, to identify accounts that may be underage. uncertain
that → use → accounts
In May, it said it would add “visual analysis” as a technique – which involves AI scanning photos and videos for “visual clues” to a user’s age. uncertain
which → say → age
Meta promised to implement child protection measures, including stronger measures to identify underage users, as part of a massive recent legal settlement in the United States. asserted
Meta → promise → States
But the company has confirmed those changes won’t be coming to Canada. asserted
changes → confirm → Canada
Google’s announcement comes after the Liberal government introduced digital safety Bill C-34. asserted
government → come → Bill
The legislation, which has not yet become law, would force social media platforms to at least temporarily block access for kids under 16. asserted
which → become → 16
The government has not yet decided how age verification would work and has said it will consult with platforms on specific methods. asserted
it → decide → methods
In December, Australia became the first country to implement age restrictions for social media. asserted
Australia → become → media
It introduced a law that requires social media platforms to take reasonable steps to keep kids under 16 from using their services. asserted
that → introduce → services
The law doesn’t include standards for age verification. asserted
law → include → verification
In the United Kingdom, legislation sets out seven separate methods of age verification for legal access to adult content. asserted
legislation → set → content
They include facial age estimation (which involves a third-party service using a live image to estimate age), a digital ID, or using a credit card, a mobile network operator or a bank to verify age. asserted
service → include → age
It also allows for the use of e-mail to estimate age through a process that looks at where an e-mail address has been used previously. asserted
address → allow → process
💬 Give feedback
🕘 History 🎫 Support