Meta’s massive child safety settlement promises to change how young people use Instagram and Facebook, but the protections have an obvious weak spot: they only work when Meta correctly knows how old a user is.
The roughly $18 billion agreement with a bipartisan group of 52 attorneys general includes strict limits for teen accounts, but Meta age verification technology will play a major role in deciding who actually receives those protections. According to Meta’s announcement of the agreement, the company plans to strengthen technology that identifies users between 13 and 17, including teens who enter an adult birthday, while improving systems designed to catch accounts belonging to children under 13.
That sounds straightforward until the technology has to make the call.
Research from the National Institute of Standards and Technology found significant room for improvement across facial age estimation systems. NIST reported that accuracy can change depending on the algorithm, a person’s age and sex, image quality, and region of birth. The agency also found no single algorithm that consistently outperformed the others. That uncertainty becomes a major issue when Meta age verification is effectively the gatekeeper for restrictions created to protect minors.
Dr. Alexis Ingber, a communications professor at Syracuse University, put the problem plainly in an interview with TechCrunch. “The design changes that were proposed in the settlement seem great on paper, but they are all reliant upon successful, effective, and unbiased age verification technology,” Dr. Ingber said. “You need to be able to identify that that’s a child versus an adult, and the way that we do that right now is not really effective… I’m very curious to see how Meta is going to move forward with trying to do that, given that by and large, these technologies have failed.”
Under Meta’s stated settlement terms, users identified as teens will face a default two-hour daily limit across Instagram and Facebook. They will also be blocked from most app activity between midnight and 6 a.m., receive fewer notifications during school hours, and see prompts during extended periods of use. Like counts will be hidden by default, and parents will receive additional controls.
The problem is that a child incorrectly classified as an adult could potentially avoid the very safeguards at the center of the agreement. An adult incorrectly classified as a child could face restrictions that were never intended for them. Guidance from the United Kingdom Information Commissioner’s Office similarly warns that age assurance systems can produce both types of incorrect outcomes and that algorithmic age estimates should be treated as estimates rather than established facts.
Meta age verification also creates another problem: proving age can require collecting more sensitive information in the name of protecting users.
“This is a classic debate around prioritizing safety over privacy,” Dr. Ingber told TechCrunch. “The perspective is, if we want to keep children safe on Meta platforms, here are all the design things we’re gonna have to do, but we’re gonna have to also do this age verification thing that’s gonna take away from people’s data privacy and will introduce these new data privacy concerns.”
Age assurance systems can use government identification, facial analysis, or other signals to estimate whether someone falls above or below a particular age threshold. Each approach raises different concerns. Facial systems can produce inaccurate estimates, while identification-based systems can require users to submit highly sensitive personal information.
Discord saw how quickly that concern can turn into backlash. According to the Associated Press, the company postponed its global age verification rollout in February 2026 after users raised privacy concerns over face scans and government identification checks. The controversy intensified following a breach involving a former outside provider that exposed government identification images belonging to thousands of Discord users.
Philip Yannella, cochair of the privacy and data security group at Blank Rome, told TechCrunch that the technology has improved and does not necessarily require companies to permanently store sensitive information.
“I think the technology has evolved quite a bit in the last few years since these concerns started coming up, and I do think there are ways that [companies] can provide or implement age assurance technologies,” Yannella said. “I’m not going to say they won’t have potential privacy issues, but they could greatly lessen the privacy issues.”
“There are ways you can verify someone’s age, create a token that would signal that this particular IP is associated with someone who’s below the age of 13, below the age of 16, whatever it is, but discard the personal information,” he said.
Still, eliminating long-term storage does not eliminate every risk. Sensitive identity information may still have to be processed or transmitted during verification. That puts Meta in the difficult position of needing a system accurate enough to enforce legally significant protections while minimizing the amount of private information users must surrender.
The financial structure makes the stakes even bigger. According to Meta, about $12.7 billion of the agreement is allocated over 10 years, while approximately $5.3 billion is conditional on TikTok and YouTube adopting specified protections and making matching payments. Reuters reported that Meta denied wrongdoing while agreeing to the settlement.
Jessica Nall, a technology litigator at Withers, told TechCrunch that the agreement “gives the AGs what they want, which is the political story that they’ve extracted a huge amount of benefit for the public from this settlement, including the platform changes,” while noting that spreading payments across a decade reduces the immediate financial impact.
“The fact that it’s being paid over 10 years really does blunt the financial impact of the large number, especially compared to Meta’s annual revenue… We’re talking about one of these tech companies where their annual revenue is larger than the GDP of most European countries.”
Nall also said avoiding a definitive legal ruling may have been important for Meta. “I think it’s wise of them to try to get out of it, especially without a legal finding against them on any of the First Amendment issues and the Section 230 issues, which are still alive,” Nall said. “That could be existential, potentially, not just for Meta but for every tech company that has a product that engages with users, which is almost all of them.”
For all of the settlement’s ambitious restrictions, Meta age verification remains the technical foundation underneath them. If the technology cannot reliably separate adults from minors, the protections can become inconsistent before screen limits, nighttime blocks or parental tools ever come into play.
“Meta has extraordinary power to create norms through its platform design,” Ingber said. “If you change the way that the platform is designed, you can set new norms. You can make things better.”
