Google is testing a broader way to identify whether users are under 18 in the U.S., using machine learning to estimate age from activity linked to Google accounts. The experiment could affect how content, ads and product features work across Google services.
What Google is testing
The company is using machine learning-powered technology to estimate a user’s age across Google accounts. According to the source article, Google may consider account-related data, including the types of details a person has searched for and categories of videos watched on YouTube.
The goal is not limited to one product. Google is experimenting with age gating for Google accounts at large in the U.S., after rolling out similar measures for YouTube earlier this week.
If the tool decides that a user is under 18, Google says the user would receive an email explaining how Google products might change for them. That notification matters because the change is not just a label in the background; it can alter access, recommendations and safety settings across several services.
What changes for users marked under 18
For accounts identified as under 18, Google plans to apply a set of restrictions and protections. These changes reach across advertising, Maps, YouTube and the Play Store.
- Google will disable the timeline in Maps.
- Google will stop personalized ads.
- Google will restrict age-restricted ad categories.
- Google will prevent access to adult-themed apps on the Play Store.
- YouTube will turn on digital well-being features, including break reminders and “go to bed” notifications when videos are watched late at night.
- Google will limit repeated access to certain kinds of content, including content that could trigger body image issues.
Taken together, these steps show how age estimation can become a product-wide control layer. A single age judgment may influence what an account can see, what ads it receives, and which prompts appear during late-night viewing.
The YouTube changes are especially notable because they combine access limits with behavioral prompts. Instead of only blocking specific material, Google is also adjusting the experience around repeated viewing and screen time.
How users can challenge the decision
Google is also offering an appeal path for people who believe their account has been incorrectly treated as underage. The source article says users can prove their age with a photo of their government ID or a selfie.
That appeal process points to a key distinction in Google’s approach: age estimation and age verification are not the same thing. Estimation uses signals to make a prediction, while verification asks the user to provide stronger proof when needed.
“Age assurance helps us ensure that adults can access the information and services they need, while also applying the right protections for our younger users. Our approach to age assurance uses a combination of age estimation and, when necessary, age verification,” Google said.
For users, the practical issue is accuracy. A mistaken under-18 classification could reduce access to certain services or features until the user appeals. At the same time, Google is presenting the system as a way to apply protections to younger users without relying only on self-declared age.
Why platforms are moving toward age gating
Google is not the only platform trying this kind of technology. The source article notes that Instagram and Roblox have also started using AI to estimate the age of their users.
The broader pressure comes from rules and proposals around age-appropriate online experiences. In the U.S., several states have existing or proposed laws around age gating. In the U.K., platforms have started age verification following the passing of the Online Safety Act.
This creates a difficult balance for large online services. They need to give adults access to information and services, while also changing the experience for younger users when certain content, ad categories or apps are involved.
Google says it has tested age-estimation technology in some markets and that it worked well, though the company did not specify which markets. The current experiment extends that idea to Google accounts in the U.S., making the account itself the center of the decision rather than treating each product separately.
What the experiment signals
The test shows how machine learning age estimation is becoming part of mainstream platform governance. Instead of waiting for a user to enter an age, platforms are increasingly looking at account behavior and content patterns to decide which protections should apply.
For Google, the experiment connects several sensitive areas: search activity, YouTube viewing, personalized ads, app access and well-being prompts. That makes the age estimate more than a compliance feature. It becomes a switch that can change the daily shape of a Google account.
The company has framed the system as age assurance, using estimation first and verification when necessary. The important question now is how well that approach works in practice, especially for users who are wrongly classified and must prove their age to restore the adult version of Google’s services.