This information was revealed by YouTube CEO Neal Mohan in his annual letter published on February 11. Google also shared more in a blog a day later.

According to Jenn Fitzpatrick, Vice President of Core Technology, the company will begin testing its machine learning-based age estimation model in the US.

The age estimation model will use available data about the user, including the websites they visit, the types of videos they watch on YouTube, and how long they have had their account open to determine age.

When Google detects that someone may be under 18, it will notify you that some settings have been changed and will provide information on how to verify age with a selfie, credit card, or ID.

on8nrttm.png
Google will use machine learning to determine users' ages, starting in the US. Photo: The Verge

Google will apply safety features to accounts under 18, including a SafeSearch filter to help remove harmful content from search results.

It also restricts content on YouTube that is inappropriate for people under 18. Google plans to roll out age-estimation technology to more countries over time.

The change comes as Google faces a slew of new regulations related to child safety online.

In addition to the Children's Online Safety Act (KOSA) and COPPA 2.0, major tech companies like Google must comply with the Children's Online Social Networking Act (KOSMA), which aims to prevent children under 13 from accessing social networks.

KOSMA proposes that online platforms should guess users' ages.

Meta has begun using AI to scan for “signals” that indicate a user may be under 18.

Google has billions of users across its platforms, and those under 18 are restricted from some services.

(According to The Verge)