California's under-16 feed ban just capped the ad-revenue engine of Meta, TikTok and Snap

California just capped the algorithmic feed for anyone under 16. On Thursday, Governor Gavin Newsom signed AB 1709, the first state law to bar social media companies from serving "addictive" features to users younger than 16 — the personalized, infinite-scroll, autoplay feed that turns attention into ad impressions. It was one of 13 child-online-safety bills signed that day, a package that also put time limits on AI chatbots for minors and raised the liability platforms face when their products harm a child.

To an investor the subject looks trivially small: a cohort, not a business. Today's signed bill even "capped" it. Whether the cap stays confined to that cohort — or reaches everyone who uses the app — is the question that decides how much of Meta's, TikTok's and Snap's ad revenue is really at risk. That is where this law bites, and the answer is more specific than the headlines.

What the statute does is narrower than it reads. It does not banish children under 16 from the platforms. They may keep an account as long as the company does not hand them the addictive feature. Under the operative text, a covered platform simply may not provide an "addictive feature" — defined as a personalized addictive feed plus autoplay — to a user under 16. Compliance therefore turns on one thing: knowing who is 16 or older before deciding whether to serve the algorithmic feed. The law routes that verification through an operating-system age signal created at device setup, with a fallback that lets a platform decline to serve the feed to anyone it cannot reasonably confirm is an adult. The enforcement teeth are real: up to $50,000 per affected minor for a knowing violation, pursued by the attorney general or a local prosecutor.

That structure is the tell. The statute is written to confine the damage. Adults keep the feed; only the under-16 cohort loses it. Nothing in AB 1709 requires Meta to strip infinite scroll and autoplay for everyone to comply. So the genuine investment question is the one the law itself poses: does the compliance burden force a costly all-users redesign, or does cheap age-gating keep the cap on minors only?

The weight of evidence points to age-gating, not a rebuild for everyone. Meta already signed an agreement in late August to pay a coalition of U.S. states and territories up to roughly $18 billion and to wall off its teenage users with daily time limits, overnight lockouts and hidden likes. Its age-assurance machinery — AI that estimates age from context and photos, plus ID checks deleted after 30 days — is exactly the tooling a minors-only cap requires. AB 1709's dependence on an OS-born age signal and a "reasonably determine" fallback pushes the same way: it builds a wall around the under-16 group rather than rebuilding the product.

So the headline is fair, but it is a teens-only cap, and that is what sizes the impact company by company. By the Harvard T.H. Chan School of Public Health's estimate, U.S. platforms took in nearly $11 billion in ad revenue from users under 18 in 2022, and the share was wildly uneven. Snapchat derived roughly 41% of its ad revenue from users under 18; TikTok around 35%. Snap's audience skews young enough that a meaningful share of its users are under 16 — losing the algorithmic feed for that cohort is not a rounding error for a company doing about $1.7 billion a quarter. For Meta, which booked more than $50 billion of revenue in a single quarter this spring, U.S. under-18s are a small slice of a giant base, and a California under-16 cap is close to immaterial to the numbers. One law, three very different exposures.

The real risks for the sector are not the law's existence but the two ways it could widen. The first is over-compliance. If OS age signals are too easy to game and one knowing violation costs $50,000 a child, the cheapest sure answer may be to withhold the personalized feed from a broader population — including adults who will not hand over paperwork. That is the self-inflicted version of the universal redesign, and it would compress time and impressions across far more than the under-16 cohort. The second is a court striking the law down entirely, which would unwind the cap in California but not the national drift that produced it.

Courts are the least-likely near-term escape hatch. The addictive-feed restriction is not new in form: 2024's SB 976 already barred personalized feeds for minors without parental consent, and last September the Ninth Circuit upheld that feed ban even as it struck down a separate likes-count provision as content-based speech. NetChoice, which lost that round, will almost certainly challenge AB 1709 on the same grounds, and the still-unresolved question of whether an algorithm is protected speech could someday win on the merits. But pending that, the feed restriction is the provision most likely to stay standing, and it now has a hard under-16 form.

The durable read is that California has legally capped the engagement engine for the cohort platforms were betting on as their future ad-viewers — a real, enforceable cap that landed this week. But it is a minors-only cap, not a strip-it-for-everyone one, and its cost is distributed unevenly: it is a headline for Snap, a footnote for Meta. The two forks that could turn this into something that re-prices the sector's growth durability — or into nothing — are whether platforms ship minors-only gating or quietly simplify for everyone, and whether a court later unwinds the law. The cap is signed. Where it ends is still being written.