In recent years, technology firms have been reluctant to constrict minors’ access to their platforms, citing technical constraints, practicality issues, overreach, and security threats. However, an increasing number of governments are now viewing these challenges as surmountable and are moving forward with stringent new age-verification requirements for social media networks, AI chatbots, and adult content providers alike.
Following Australia’s notable prohibition on adolescent social media accounts three months ago, regulatory bodies across Europe, Brazil, and several U.S. states are seeking to replicate the move. High-profile political figures such as California Governor Gavin Newsom and former President Donald Trump have also reportedly shown interest in implementing age restrictions.
The driving forces behind these regulatory changes are mounting anxieties over online abuse and adolescent mental health issues, recent outrage over AI-generated explicit images of children, and a growing faith in the potential of ‘age assurance’ software. Such software can estimate an individual’s age based on facial analysis, parental permission, ID checks, and other digital hints.
Recent strides in artificial intelligence have enhanced the efficiency and reduced the cost of age-verification tools. Social media enterprises can now often accurately determine a user’s age group using digital indicators such as the account creation date and the type of content viewed. Meanwhile, an emerging industry of age-verification vendors offers added layers of checks through automated tools like facial scans and machine-based analysis of government IDs.
App store platforms like Apple and Google have also introduced tools that enable parents to communicate their child’s age range to app developers. This overall improvement in technology and the resultant drop in verification costs have expanded the scope of its application.
Age-verification vendors usually charge less than a dollar per check for basic machine-only age-assurance tools. Traditional methods that were standard a decade ago, such as human confirmation and triangulation of personal data, are still available at a premium but are needed less frequently. Furthermore, independent evaluations show progress in the accuracy and precision of these tools.
For example, face-scanning software was off in age estimations by an average of 4.1 years in 2014, while by 2024, this average has dropped to 2.5 years. The latest models from vendors like Yoti and Persona boast of an average error of less than two years for the age range of 13 to 18.
Despite these advancements, there are still limitations. The systems face difficulties with certain skin tones and older phones. Also, they might fail to catch attempts by youngsters to appear older than they are. However, executives believe that facial age estimation can offer a digital equivalent of offline age screening.
Adding to this, social media services generally require fewer facial scans and ID checks than adult content or gambling sites due to the abundance of personal information they have on users. This allows them to depend more on an age-verification method called ‘inference’.
While it’s too early to assess the full impact of Australia’s teen ban, preliminary results are promising. According to Australia’s eSafety Commissioner, companies have blocked 4.7 million suspected underage accounts since the law came into effect. Regulators in other countries are closely monitoring these developments as they consider implementing similar measures.
What are some of the driving forces behind the implementation of age verification measures?
Concerns over online abuse, adolescent mental health issues, and the spread of AI-generated explicit child images have all contributed to the push for stricter age verification measures.
How have technological advancements influenced the age assurance market?
Advancements in artificial intelligence have improved the efficiency and lowered the cost of age-verification tools. This has broadened the scope of their application beyond high-value transactions.
What limitations do current age-verification systems face?
Current age-verification systems can struggle with certain skin tones and older phones. They may also fail to catch attempts by youngsters to appear older than they are. These challenges highlight the need for ongoing refinement and improvement in this technology.
