Google uses machine learning to estimate users' age and blocks them from restricted content and ads

Machine Learning


This week, Google will begin using machine learning to estimate users' age to better tailor the advertising experience for minors, the company said in a blog post Wednesday.

This technology helps you determine if a user is less than 18 using behaviors such as search queries and the type of video consumed on YouTube.

If a user is flagged as likely to be under the age of 18, the system will notify the user and automatically implement guardrails across Google's products. Alcohol, gambling, weight loss, high fat and sugar foods and drinks.

Google also identifies minors as they are prohibited from accessing restricted apps to adult users of the Google Play Store and are automatically selected for YouTube's digital well-being program, including content protection, restricting repeat views of several types of videos, and reminders to take a break from the platform. Google Maps Timeline settings maintain a timeline record of the locations users have visited, but are turned off for users who are presumed under the age of 18.

Google will be testing a new feature called “Age Assurance” for some US sign-out users over the coming weeks.

The changes were compiled into a blog post as part of the company's efforts to “more protect young people when using Google products,” and were shared by email with Google ad customers on Wednesday afternoon.

The development follows a February announcement from YouTube CEO Neal Mohan, who outlined this year's plan to use machine learning to expand advertising protection for minors.

Last fall, Google accused Google of intentionally targeting teens on YouTube for violating platform policies, Adweek previously reported.

The deployment of Google's “age guarantee” tool comes just a few months after Meta introduced similar products to Instagram. This helps to ensure that minors use protected “teen accounts” for users under the age of 18, rather than an unlimited version of APP.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *