Select Language:
For years, technology companies resisted pressure from child safety advocates, arguing that technical limitations made it impractical, overly broad, or risky to restrict teenage access to their services. They claimed that implementing effective restrictions would be either unfeasible or pose security concerns.
Recently, however, more governments are beginning to view these obstacles as manageable. Several are moving forward with strict new age verification measures for social media platforms, AI chatbots, and adult content sites alike.
Three months after Australia introduced a groundbreaking ban on social media accounts for those under 16, regulators in Europe, Brazil, and a few U.S. states are planning to follow suit. California Governor Gavin Newsom, considered a potential Democratic presidential candidate in 2028, publicly supported the initiative last month. Meanwhile, former President Donald Trump, a Republican, is reportedly “interested” in age restrictions, according to his daughter-in-law.
Heightening the pressure are mounting concerns over online abuse, teen mental health issues, and recent outrage regarding AI-generated images depicting child sexual abuse. Confidence in “age assurance” technologies—software that estimates ages through facial analysis, parental approval, ID verification, and other digital data—has increased among developers and regulators. These tools aim to confidently determine a person’s approximate age without overly intrusive measures.
Advancements in artificial intelligence have significantly improved the accuracy of these tools and reduced their costs. Interviews with more than a dozen regulators, child safety advocates, independent researchers, and vendors that provide age verification solutions for major tech giants like TikTok, Meta, and OpenAI confirm this progress.
“The market for age assurance has matured considerably over the last few years,” said Ariel Fox Johnson, senior adviser at Common Sense Media, a children’s online safety organization. She highlighted improvements in technology, along with the development of industry standards, certification processes, and collaborative groups that assess the efficacy of various tools.
Today, social media companies can often accurately estimate a user’s age group by analyzing digital footprints such as the account creation date or the content viewed. A growing ecosystem of vendors—like Yoti, k-ID, and Persona—offers layered verification methods, including facial recognition, ID scans, and automated analysis of government-issued documents.
At app stores, developers can specify age ranges for their apps, with Apple and Google providing tools for parents to detail their children’s age brackets to developers.
“The technology has definitely gotten better—not just for age verification, but for overall identity validation,” said Merritt Maxim, Vice President at Forrester Research. “That has driven down costs, making it feasible to verify identity across a broad spectrum of online interactions, not just high-value transactions.”
Verification providers typically charge under $1 per check for basic, machine-only methods, with costs as low as a few cents for high-volume scenarios. Traditional, manual checks—such as human confirmation and triangulation of personal data—remain available but are now less common due to their higher costs and resource requirements.
Independent evaluations substantiate these claims. A study led by the U.S. National Institute of Standards and Technology (NIST) shows that face-scanning software from companies including Yoti—used by TikTok, Meta, Instagram, and Threads—reduced age estimation errors from an average of 4.1 years in 2014 to about 2.5 years in 2024.
Yoti reports that its latest facial analysis model, set for release in April, surpasses previous versions, estimating ages within about one year for teenagers aged 14 to 18. Similarly, Persona, used by OpenAI and Reddit, claims an error margin of roughly 1.8 years for ages 13 to 17.
A report commissioned by the Australian government last year reinforced this trend, noting that photo-based age estimation can be quite accurate but has limitations—particularly for users within three years of the legal age threshold of 16. The report recommends supplementing facial analysis with additional verification methods, such as ID checks or parental consent, especially for borderline cases.
Facial verification systems face challenges with various skin tones, image quality, and privacy-preserving on-device processing—where data analysis occurs on a user’s device rather than cloud servers. Such factors can impact accuracy, especially when attempts are made to appear older through masks, makeup, fake facial hair, or by scanning toys’ faces instead of their own.
Nevertheless, executives compare facial age estimation to the age screening done at liquor stores or bars in physical retail environments, where individuals are challenged to provide ID if they appear underage.
“If you look young, you might be asked to show your ID,” explained Robin Tombs, CEO of Yoti. He noted that social media platforms generally require fewer face scans and ID checks than adult content or gambling sites because they already possess extensive user data. These platforms often rely more on “inference”—analyzing online activity, financial data, and other signals—to confirm age and meet regulatory standards.
The ten social media firms subject to Australia’s teen account ban declined requests from Reuters to share data on their particular tools’ effectiveness.
In Australia, authorities plan to monitor the ban’s impact over two years, with results anticipated later this year. Since December, the law has led to the suspension of approximately 4.7 million suspected underage accounts, although some may be inactive or mistaken identifications. Meta reports removing about 550,000 suspected underage accounts on Instagram, Facebook, and Threads, while Snapchat has taken down around 415,000.
Other regions are observing these developments carefully. European officials, including European Commission President Ursula von der Leyen, are expected to discuss age verification during her upcoming visit to Canberra. The UK, which already mandates age verification for pornography websites and is considering tighter child safety laws for social media and AI, is exchanging insights with Australian policymakers.
Initial results from Australia’s enforcement should be interpreted cautiously, as companies often played it safe—asking vendors to disable stricter controls, according to Iain Corby of the Age Verification Providers Association. Corby suggests that many firms are testing the waters to see how much regulators will tolerate, rather than fully embracing rigorous enforcement.
“Companies are worried this could become a global policy, and they’re mainly trying to see what they can get away with,” Corby explained.
ChatGPT
Perplexity
Gemini AI
Grok AI
