New York Attorney General Letitia James and Governor Kathy Hochul released the final rules for the SAFE for Kids Act on July 28, 2026. The rules say social media platforms must check a user's age before letting them see algorithmic feeds — the streams of posts an app picks for you based on what it thinks you'll like — and before sending notifications late at night. (New York AG press release)

The law was signed by Hochul in June 2024. SAFE stands for "Stop Addictive Feeds Exploitation." (Verge)

The final rules are largely similar to an earlier version proposed in September 2025. The public had a chance to comment on the proposal through December 1, 2025. (Verge)

Several specific requirements are now set. Age-check methods must catch people trying to get around them 98 percent of the time. Platforms can accept a government ID but must also offer at least one other way to verify age. Any information collected to check age must be deleted or made anonymous immediately afterward. (Verge)

The law applies to social media platforms where users create content and spend at least 20 percent of their time on algorithmic feeds. That threshold is meant to cover the big platforms where recommended content drives engagement while excluding apps where it's not a central feature. (Verge)

Two features trigger the age requirement: algorithmic feeds and notifications sent between midnight and 6 AM. Anyone under 18 must get a parent's permission to use these features, and the platform must notify the parent of the request. Both parents and minors can take back that permission at any time. (Verge)

The rules include some limits on how the consent process works. Platforms cannot block minors from using other parts of the app just because a parent hasn't agreed to the age-restricted features. Platforms also don't have to show parents their child's search history or content topics as part of the consent process. (Verge)

Breaking the rules can cost up to $5,000 per violation. The law takes effect January 25, 2027, giving platforms about 18 months to build the required systems. (Verge)

The broader context here is the legal landscape. The US Supreme Court recently upheld a Texas law requiring age verification on adult websites. That decision changed how courts view online age checks. First Amendment arguments that might once have been strong against these requirements are now harder to make. Whether the SAFE for Kids Act survives its own legal challenges will depend on how courts distinguish between age-gating adult content and age-gating algorithmic feeds and notification timing.

For the companies building these systems, the 98 percent accuracy standard is the hardest part. Government ID checks can reach that level, but having to offer a non-ID alternative pushes platforms toward technologies like estimating age from a photo or using third-party services, each with accuracy and privacy trade-offs. The rule that age data must be deleted right away means platforms can't keep it for other purposes like training AI or catching fraud. And the consent system, which gates specific features rather than whole apps and can be withdrawn at any time, is more complex than what most platforms currently have in place.

The SAFE for Kids Act is part of a larger New York effort that also includes the NY Child Data Protection Act. Together, these laws require platforms to build child-safety features that go well beyond what federal law currently demands.

For an industry built on making sign-up as easy as possible, these requirements are not simple. But the 18-month timeline and the clear scope of the rules give companies room to work. The platforms that start building early, rather than waiting until the last minute, will be better prepared for what looks like an inevitable wave of similar state laws.