The agreement, reached in August by Meta and 47 states, Washington, D.C., and three U.S. territories, bars minors from using certain “cosmetic procedure filters” that simulate extreme changes, such as plastic surgery or heavy makeup. Meta has said it plans to enforce the restriction by early 2027. State attorneys general set a two-month window to give the company guidance on which filters should be prohibited.

The settlement does not ban every effect. Playful filters that turn users into animals or fantasy characters will remain available, as will subtler features such as smoothing skin or brightening eyes. Those less dramatic edits are a particular concern for Mia Lerner, a 17-year-old California student and youth body-image advocate.

Lerner said experimenting with beauty filters contributed to her eating disorder and damaged her self-image. She has used artificial intelligence to alter her nose in photos, she said, but still has to live with the face she sees in the mirror. She supports the restrictions, calling them “a very important step in the right direction.”

Lerner said obviously manipulated images can be easier to recognize as unrealistic than videos featuring slight edits, which she finds more likely to trigger comparisons. She also urged technology companies and policymakers to hear directly from teenagers as they decide what safeguards will help.

Meta has previously prohibited filters that explicitly promote cosmetic surgery, including ones that add pre-surgery markings to a user’s face. It has also ended its platform for outside developers of filters, introduced labels for AI-generated content and deployed automated detection systems. But a federal judge wrote in a summer ruling rejecting Meta’s defense that harmful aesthetic features and designs encouraging prolonged use continued to evade safeguards.

The rules differ across social media platforms. TikTok restricts beauty filters for users under 18, while Snapchat does not limit teens’ access to filters. YouTube said it does not offer filters, but users can upload images and videos edited with outside tools to YouTube and other platforms.

Defining a prohibited filter may prove difficult as ideals change. Florencia Solari, a creative technologist who has worked on augmented-reality and AI experiences, said the look once known as “Instagram face”—with a small nose, prominent cheekbones and full lips—is already giving way to other trends, including extreme thinness.

That change makes any fixed technical definition vulnerable to becoming outdated, said Kelly Frasier, co-author of a study on social media and adolescent body-image concerns. Frasier argued that lawmakers should consider broader rules for AI use by minors, rather than focus only on filters that alter a person’s face.

Legal questions also remain if states set conflicting standards. David Greene, senior counsel at the Electronic Frontier Foundation, asked what would happen if one state permitted a filter that another considered unacceptable. Attorneys general in California, Colorado, Kentucky and New Jersey—the four states that led the trial—either declined to comment on guidance or did not respond to requests for updates.

The European Union is taking a broader approach in a proposed youth social-media law: it would prohibit all beauty filters for minors, not only those considered extreme. European Commission spokesperson Thomas Regnier said EU officials are watching how the U.S. settlement is implemented.

For Lerner, the debate should include young people who use the platforms. She said companies should ask teens whether the filters targeted by the rules are the ones causing harm—and whether other features deserve more attention.