Google expands age assurance technology to Canada as Ottawa proposes under-16 social media ban Canada has introduced legislation that would ban children under 16 from social media, while technology companies are expanding systems that estimate or verify users' ages. Google and Meta use different methods, including account activity, artificial intelligence and proof-of-age checks. OTTAWA - The federal government has introduced digital safety legislation that would introduce a ban on social media for kids under 16. Big tech companies are already using age assurance technology to estimate users' age, with Google now expanding the technology to Canada. Age assurance is the umbrella term used for efforts by tech platforms to determine the age of a user. It includes age declarations, such as the click-to-verify-you-are-18 boxes that internet users have been encountering for a long time. More recently, technology companies have been incorporating age estimation. That includes the use of artificial intelligence to estimate a user's age based on factors such as their online behaviour or the images they post.Age verification is any method that requires a user to prove their age. The use of government identification is one example, although companies are also exploring checks involving selfies, credit cards and other methods. Google said last year it would begin rolling out age assurance in the United States. On Monday, it announced it would expand the technology to Canada, as governments and platforms face increasing pressure to protect young users online.Google's approach looks at factors such as a user's search queries and the types of videos they have watched on YouTube. The company says it will not collect any additional information for age assurance. Users who are identified as underage by Google have additional protections placed on their accounts. Those measures include tools that remind users to take a break and safeguards on content recommendations.The protections also include disabling personalized advertising and restricting access to apps that are adults-only. The measures are intended to adjust the experience for accounts that Google believes belong to younger users. If the technology mistakenly identifies an adult as being under 18, the user will be able to prove their age through tools such as a government ID, a selfie or a credit card check.Google says that users who provide a government ID will have the date of birth stored in their account, while the identification document itself will be deleted after a short period of time. The company says it will use the service PrivateID to analyze selfies. PrivateID will tell Google whether it believes the user is of age, but Google says the service does not provide it with the selfie or the precise age of the user.Google also says PrivateID does not retain the image. The company's system therefore combines estimates based on account activity with an opportunity for users to challenge an assessment through an age verification process. Meta uses a different approach. If it suspects the person behind an account is under 18, it places the account in a teen account, with settings that provide additional protections.Those settings limit who can contact the user and the content they can view. Meta also uses artificial intelligence to remove individuals it believes are under 13 from its services entirely. Meta has said its artificial intelligence analyzes entire profiles for contextual clues that could indicate an account belongs to someone underage. Examples include birthday celebrations or mentions of school grades.That work includes looking at posts, comments, biographies and captions. The company's system is designed to assess the wider context of an account rather than relying only on the age a user entered when signing up. If Meta determines an account may be underage, it can be deactivated. The account holder must then provide proof of age through Meta's age verification process to prevent the account from being deleted.The company said in May it would also begin using visual analysis. That technology allows its artificial intelligence to scan photos and videos for visual clues about a person's age. Meta said the visual analysis does not identify the person in the image and does not use facial recognition. The company has not described the system as a way to establish a person's exact age.The systems being introduced by Google and Meta illustrate the range of methods covered by the term age assurance. Some methods rely on information already associated with an account, while others ask users to submit additional evidence. The technology is being introduced as lawmakers consider stricter online safety requirements for children and teenagers. In Canada, the proposed under-16 social media ban would add to the pressure on platforms to determine which users are minors.Age declarations remain the simplest form of age assurance, but they depend on users providing accurate information. Age estimation and age verification are intended to offer platforms additional ways to identify accounts that may not match the age supplied at registration. The approaches also raise questions about privacy, accuracy and what happens when a system makes a mistake.Google says it will not collect additional information for its age assurance process, while both companies describe safeguards intended to limit the handling of sensitive material. For users identified as children or teenagers, the result may be restrictions on advertising, recommendations, contact from other accounts and access to adult-only applications. Adults who are incorrectly classified can use available proof-of-age tools to seek a correction. Anja Karadeglija, The Canadian Press