Meta offers new compliance numbers on Australia SMMA, deactivates 750K accounts

Meta says it has taken down more than 750,000 accounts believed to belong to Australian users under 16 years of age, in the wake of Australia’s landmark regulation putting age assurance requirements on large social media platforms.

According to Reuters, the company claims to have deactivated 462,000 suspect Instagram accounts and 294,000 suspect Facebook accounts from December 2025, when the law went into effect, to June 2026. That’s an increase of 331,000 Instagram accounts and 173,000 Facebook accounts it reported deactivated in January 2026.

The company is apparently eager to show the efforts it has made to comply with Australia’s law, as tech regulator eSafety weighs legal action against sites it says have violated it, and develops a broader legal framework in the form of a digital duty of care. The agency has also introduced a law to double the maximum penalty for noncompliance to $99 million ($69.75 million).

A statement from Meta maintains that it shares the Australian Government’s goal of “ensuring young people have safe, age-appropriate experiences online.”

“Enforcement is ongoing, and these ​numbers will continue to grow. We are meeting our obligations under the law.”

Among social media companies the law covers, Meta has released the broadest date range of compliance data. Nonetheless, other big players are implicated in the discussion, and TikTok, Google and Snap are joining Meta in providing evidence to a parliamentary inquiry about the changes.

Back-and-forth makes true accounting difficult

In setting a precedent that has quickly found favor among regulators globally, Australia has seen the broadest range of how platforms respond – and the most kerfuffles over legal details. Per Reuters, “Australia has accused the platforms of intentionally setting the ban ‌up ⁠to fail.” Findings that many kids under 16 remain on social media are framed as a failure of legislation; conversely, numbers show how weak the implementation of highly effective age assurance has been on the platform side.

Meta pushes its own variety of age checks based on user content. The social media firm’s preferred internal AI model analyzes “contextual clues that an account ⁠may belong ​to someone under 16, such as birthday celebrations or mentions of school ​grades” – i.e., user data – to find and police underage accounts. While this method is often referred to as age inference, the UK Information Commissioner’s Office (ICO) classifies it as “profiling,” and says it doesn’t cut the mustard as reliable age assurance tech.

Meta has long declared itself opposed to age assurance laws imposed at the platform level, and its promises to ramp up enforcement should be taken with a grain of salt. Its efforts continue to undermine trust in proven age assurance methods and in the Australian regulatory scheme, and any compliance it boasts of has surely been achieved under duress.

Article Topics

age verification  |  Australia age verification  |  biometrics  |  eSafety Commissioner  |  facial age estimation (FAE)  |  Meta  |  social media