Secure Age Verification Laws: AUD 99M Fines Can't Fix a Face Guess

Quick answer
Can facial age estimation get your age wrong?
Yes. Facial age estimation is a statistical guess, not a record of your birthday. The program returns an estimated age with a confidence score, and research found higher error rates for darker-skinned people and for women. A wrong result should be treated as a clue and be easy to challenge with an ID.
Picture a 41-year-old locked out of her own social media account at 11pm because a camera guessed she was 15. No human looked at it. No one will say why. That is the quiet worry underneath this week's age verification headlines, and it is a fair worry.
Age verification is growing fast, Australia just doubled its top fine to AUD 99 million, and a face scan is still only a guess, so you deserve a human way to challenge a wrong answer.
Age verification is getting serious: why Australia doubled its fines to AUD 99 million
On June 29, 2026, Australia raised the maximum penalty for platforms that keep failing its social media minimum-age law. The old ceiling was AUD 49.5 million. The new one is AUD 99 million. The details were laid out by Shattered.io and also covered by The Next Web.
A big number grabs attention. But the number isn't the real story. The real story is what regulators are now asking for: proof.
What Australian age verification enforcement now demands
Under the old approach, a company could say, "Yes, we have an age check." Done. Now Australia's eSafety Commissioner (the government's online safety watchdog) can demand documents and evidence. Not just from the platforms. Also from the companies that run the age checks and from app stores.
Why the shift? Research found that about two-thirds of underage users got in anyway. Some simply typed in an older birthday. Others uploaded a selfie that the platform accepted as proof. So a quick age check that looks good on paper can fail in real life, and the regulators noticed.
What does online age verification in Australia mean for the rest of us?
Short answer: more faces getting scanned, by more companies, with more riding on the result. Australia's rule is aimed at kids, but a system that guesses a child's age also guesses yours. Adults get swept up too. Every adult who gets wrongly flagged becomes a customer service problem, and sometimes worse. This article is part of a series, start with Minimum Act Score Colleges Vary Eu Social Media Age Is 15.
Here's the catch the industry keeps raising. Industry groups argue that flawless age assurance (proving someone's age with confidence) doesn't exist unless people upload sensitive biometric data (your face, voice, fingerprints, the body stuff that's uniquely you) or a government ID. That is a steep price for an adult who just wants to scroll.
Facial age estimation: a guess, not a birth certificate
Many age checks work by facial age estimation. A program looks at your face and guesses a number. It doesn't know your birthday. It has a hunch, plus a confidence score (a number showing how sure the program is). Think of a carnival worker guessing your weight. Sometimes spot on. Sometimes embarrassingly off.
Facial matching is a probabilistic tool masquerading as a certainty.
Expert analysis prepared for this story, CaraComp research notes
Probabilistic is a fancy word for "it deals in odds." Odds are fine for suggesting a next step. Odds are not fine as the last word on whether you keep your account, your access, or your freedom.
Age verification and wrongful arrests: the same flaw wearing a different uniform
This week's other big face story was about police. England's police inspectorate (HM Inspectorate of Constabulary and Fire and Rescue Services) said there are not enough safeguards to be sure facial recognition programs are free of bias against particular groups. It also warned that face matches have been treated like probable cause (a legal reason to make an arrest) even after officials promised that wouldn't happen.
Put the two stories side by side. A platform guesses your age from your face. A police system guesses your identity from your face. Different stakes, same weakness. The program returns an answer that sounds sure, and a person downstream treats it as fact.
Researchers at the Georgetown Law Center on Privacy & Technology have a pointed title for this problem: "A Forensic Without the Science." And a piece on Security Boulevard draws the line that matters: a match should be a lead to check, not proof to act on. Previously in this series: Age Verification Face Scan 2 Year Miss Lets 16 Pass As 18.
| Question | A face-based guess | What real proof looks like |
|---|---|---|
| Age verification | Facial age estimation: a program estimates a number from your face, with a confidence score | Identity verification: a checked ID or a record a person can review |
| Who is held accountable | Often unclear until a regulator steps in, as with the new AUD 99 million fines | Documents and evidence the company must produce on demand |
| Wrongful arrests | A face match treated as probable cause | A lead that is checked against other evidence first |
| Who gets hurt by mistakes | Women and darker-skinned people, per research on error rates | Fewer people, because a human catches errors early |
Do face-based systems get some people wrong more often?
Yes. Research on commercial facial analysis systems found much higher error rates for darker-skinned people and for women than for light-skinned men. So a mistaken age check or a mistaken match is not spread evenly. Some people will be told "you're wrong" far more often than others, and they'll have the hardest time proving otherwise.
Why Age Verification Matters Beyond Teens
- Your access - a wrong guess can lock an adult out of an account, with no obvious way back in.
- Your privacy - stricter age verification laws push companies to collect more verification data, such as your face or ID.
- Your fairness - error rates are not the same for everyone, so mistakes pile up on certain groups.
- Your freedom - the same shaky logic, treated as certain, has fed wrongful arrests.
Will bigger age verification fines make face checks more accurate?
Probably not by themselves. That is the fair criticism. Higher fines don't fix an algorithm. They create pressure, and pressure can push a company toward collecting more data or buying from whichever vendor sells the best story. Australia's answer is to demand evidence and audits, not just penalties. That is smarter than fines alone. But nobody has yet agreed on what "good enough" accuracy looks like when a decision lands in court or on your account.
Here is my honest take. Protecting kids online is a real goal, and nobody serious disputes it. But a system that protects some people by quietly misjudging others needs a front door for complaints. Not a hidden email address. A real appeal, with a human on the other end.
One thing you can do tonight
If you've ever wondered whether a photo, a profile, or a face is really what it claims to be, that is exactly the question this kind of technology exists to answer. The trick is treating the answer as a clue. So if a platform or service gets your age or identity wrong, do three things. Ask whether a person reviewed it. Ask whether you can submit another proof, like an ID or, where offered, a credit card check. Save screenshots of what it told you and when. That small paper trail (a record of what happened and when) is often what turns a "computer says no" into a fix.
Age verification is becoming normal, and Australia's AUD 99 million fines show regulators want proof, not promises. But a face scan only supplies a facial age estimation, so any wrong answer needs a human appeal.
Recap: Face-based age checks are spreading, and Australia now fines weak ones up to AUD 99 million. A face gives a guess, not the truth, so people need a real way to challenge mistakes.
So here's the question worth carrying to bed. If a camera decided you were the wrong age tomorrow, who exactly would you call? If the platform can't answer that, the fine isn't the biggest problem. Up next: Age Verification Aud 99m Fines Cant Fix A Face Scan Guess.
Age Verification: Frequently Asked Questions
What is age verification?
Age verification is any way a website or app tries to confirm how old you are. Some use a typed birthday. Some check a government ID, a form of identity verification. Some use a camera that guesses your age from your face. The strength of the check matters, because a typed birthday is easy to fake and a face guess can be wrong. Australia is now demanding evidence that these checks work.
What is the AUD 99 million penalty in Australia?
It is the new maximum fine for platforms that systematically break Australia's social media minimum-age rule. It took effect June 29, 2026, doubling the earlier AUD 49.5 million ceiling. It is a ceiling, not a standard bill, but it signals that regulators expect proof that age verification systems genuinely work rather than a company's word.
How does age verification in Australia affect adults?
Anyone using a covered platform may be asked to prove their age, which can mean a face scan or an ID upload. Industry groups warn that stronger age verification methods can push users to hand over sensitive biometric data and weaken their privacy. If a check gets an adult wrong, that person needs a clear path to a human review and a second form of proof.
Can facial age estimation get my age wrong?
Yes. Facial age estimation produces a guess with a confidence score, not a fact from your birth record. Research on commercial facial analysis found higher error rates for darker-skinned people and for women compared with light-skinned men. That is why a wrong result should be easy to challenge, and why a platform should offer another way to prove your age.
Why are age checks linked to wrongful arrests?
They aren't the same technology use, but they share a flaw: treating a face-based guess as certain. England's police inspectorate warned that facial recognition has been treated like probable cause despite promises it would not be. A match should be a lead for investigators to check, never proof on its own.
What should I do if an age check locks me out?
Ask whether a person reviewed the decision, and ask for another way to prove your age, such as an ID. Take screenshots showing what the platform said and when. Keep every message and any information the platform gave you about the decision. A clear record helps if you escalate to the company or a regulator, and it shows you tried the normal route first before age verification became a dispute.
Verification methods: how the common options differ
Not every age verification system works the same way, and the differences matter. The common verification methods are a typed birthday, a checked ID, a credit card check, and a face-based guess. Each one trades convenience against accuracy and against how much personal information the user must hand over.
A typed birthday is quick but easy to fake, which is how many minors online got past the early checks. An ID check is stronger, but it asks the user to share sensitive data. A face-based guess asks for no document, yet it only estimates user age instead of proving it.
Age assurance in plain English
Age assurance is the wider family of tools that try to work out whether someone meets an age requirement. It covers everything from asking for a birthday to estimating age from a face. Age verification is the stricter end of that family, where the goal is to confirm age against a trusted record.
Good age assurance should be proportionate. That means the strength of the check should match the risk of the age-restricted content behind it. A site hosting pornographic material, for example, reasonably faces a tougher age requirement than a general video app.
Age gating is the simple idea of putting a barrier in front of content until a visitor clears it. The barrier can be weak or strong. A weak gate invites minors to click past it, while a strong gate can collect more data than a digital service really needs.
Protecting children without over-collecting data
Protecting children online is the stated reason for these laws, and it is a fair goal. Still, the tools used to protect minors also touch adults. Every online age check is run on individuals, so a method that guesses wrong will hit grown-ups as well as kids.
That is why privacy matters as much as accuracy. A company should collect only what it needs to confirm the user's age, use that data for that purpose, and delete it afterward. Checks that can be done privately, without building a permanent file on a face or an ID, are better for everyone.
Regulators asking for evidence is a step in the right direction, because it lets outsiders test whether a check works. Individuals should also have a plain way to contest a result. When a digital check labels a 41-year-old as a teenager, a person with authority should be able to fix it quickly.
For readers weighing a platform, a few questions help. Does it explain how it checks age? Does it offer more than one way to prove user age? Does it say what happens to your information afterward? Clear answers suggest the company takes both children's safety and your privacy seriously.
The bottom line is simple. Age verification can be useful, but it works best when it is honest about what it can prove. A digital guess about age is a starting point, and the people it affects deserve a human review of any result that feels wrong.
Ready for forensic-grade facial comparison?
Full forensic reports with detailed similarity scoring. Results in seconds.
Run My First SearchMore News
Biometric Camera: A Facial Match Is a Lead, Not Proof
A commissioner says Scotland is in a "perilous position" over police facial recognition. Here is why a face match should be a lead to check, never proof, and what checks should come first.
ai-regulationInstagram Age Verification: AI Can Lock Adults in Teen Mode
Instagram's AI now guesses ages worldwide, and Meta has not published how often it gets them wrong. Here is what to do if the guess lands on you or your kid.
privacyMinimum ACT Score: Colleges Vary, EU Social Media Age Is 15
A reported EU law would set a minimum age of 15 for social media and require real age checks. Safer for kids is the goal. What happens to your family's data is the open question.
