Google and YouTube on November 4, 2026 sued California Attorney General Rob Bonta in federal court in San Jose, asking a judge to stop the state from enforcing a new law that bars online platforms from giving users under 16 personalized feeds or automatically playing the next video - a ban that, according to the complaint, applies even when a parent consents. The 28-page complaint, case 5:26-cv-13204-SVK in the US District Court for the Northern District of California, targets three provisions of Assembly Bill 1709, which takes effect on January 1, 2027 and carries civil penalties of up to $50,000 per affected minor.

The filing is Google's second attempt in a year to keep California from limiting how YouTube's feeds work for young users. Its first, against an earlier state law, failed at the preliminary stage in August. This time the company leads with a different argument: that federal law, specifically Section 230 of the Communications Decency Act of 1996, preempts the state's rules because choosing which video to show next is the work of a publisher. The complaint is signed by Elizabeth B. Prelogar of Cooley LLP, a former US Solicitor General. Its allegations have not been tested in court.

In Short

California passed a law that stops apps and websites like YouTube from choosing videos for kids under 16 based on what they watched before, and from starting the next video automatically. Google, which owns YouTube, has asked a federal judge to stop the state from using that law against YouTube, arguing that picking which videos to show is what a publisher does and is protected by federal law and the US Constitution. If no judge steps in, YouTube would have to change how it works for every Californian under 16 by January 1, 2027, or risk fines of up to $50,000 for each young user affected.

What Google wants from the court

The plaintiffs are Google LLC, a Delaware company wholly owned by Alphabet, and YouTube, LLC, a subsidiary of which Google is the sole member. The defendant is Bonta, sued in his official capacity because the Act gives the attorney general enforcement authority under Section 22686(a) of the California Business and Professions Code.

The complaint sets out four counts. Count I, brought under Ex parte Young and 42 U.S.C. Section 1983, claims that Section 230 preempts the law's ban on personalized feeds, its ban on autoplay and the related requirement to delete accounts. Count II claims the same provisions violate the First Amendment, as incorporated by the Fourteenth Amendment. Counts III and IV ask for equitable relief and a declaratory judgment.

The scope matters. The First Amendment count is limited to personalized feeds; autoplay is challenged only under Section 230. That split runs through the prayer for relief, which asks for a declaration that Section 230 preempts Sections 22683(a)(1), 22683(b) and 22684(b)(1) "as applied to YouTube's personalized feeds and autoplay," and a separate declaration that the same sections violate the First Amendment "as applied to YouTube's personalized feeds." Google also asks for a preliminary and then permanent injunction barring Bonta and anyone acting under his direction from enforcing those sections against Google or YouTube, LLC, plus attorneys' fees under 42 U.S.C. Section 1988(b).

This is an as-applied challenge. Google is not asking the court to strike down AB 1709 for every platform, only to stop its enforcement against YouTube.

On standing, the complaint is careful with its words. YouTube "has reason to believe that it is a 'covered platform'" under the Act, it says, and would face "a serious risk of an enforcement action" if it did not comply. It points out that in the earlier case, Bonta "refused to stay enforcement of the law pending YouTube's appeal." The plaintiffs also claim standing to assert the First Amendment rights of YouTube's under-16 users, citing Virginia v. American Booksellers Association (1988) and Ninth Circuit authority allowing platforms to raise their users' rights.

What AB 1709 requires

According to the complaint, California enacted AB 1709 in September 2026. Governor Gavin Newsom signed it on September 10, after the legislature passed it without a no vote on August 31.

The law regulates platforms that offer what it calls "addictive" features. Its definition of an "addictive feed" covers "an internet website ... in which multiple pieces of media generated or shared by users are ... recommended, selected, or prioritized for display to a user based ... on information provided by the user, or otherwise associated with the user or the user's device," with narrow exceptions. Autoplay is named as a second addictive feature. The complaint notes that the Act does not define autoplay at all.

Three provisions are in dispute:

  • Section 22683(a)(1), the core ban: "A covered platform shall not provide an addictive feature to a user who is under 16 years of age."
  • Section 22683(b), which requires platforms to "implement reasonable measures to ensure that users under 16 years of age are not provided any addictive feature on the covered platform."
  • Section 22684(b)(1), which requires platforms to "delete the account of a user under 16 years of age and any personal information associated with the user of the account" to the extent that user is provided with a personalized feed or autoplay.

Who is covered

A "covered platform" is "[a]n internet website, online service, online application, or mobile application, including, but not limited to, a social media platform" that offers enumerated features, including personalized feeds and autoplay, "as a significant part of the service." The feed definition turns on media "generated or shared by users," and the complaint argues that this leaves video services such as Netflix and Hulu untouched, a point it uses in its constitutional attack.

Penalties and timing

The Act goes into effect on January 1, 2027, and threatens civil penalties of up to $50,000 "per affected minor," according to the complaint. That is the ceiling for a knowing violation; a negligent violation carries up to $25,000 per affected minor, and only the attorney general or a local public prosecutor can enforce the law.

The per-minor structure features heavily in Google's argument on harm. "YouTube cannot recover the costs of compliance from the State," the complaint says. "And because penalties attach separately to each affected minor, YouTube has no realistic option of violating the Act once and raising Section 230 as a defense."

The Section 230 argument

Section 230(c)(1) states that no provider of an interactive computer service "shall be treated as the publisher or speaker of any information provided by another information content provider." Section 230(e)(3) bars liability "under any State or local law that is inconsistent with" that rule. Google walks through the three-part test from Barnes v. Yahoo! (9th Cir. 2009) and claims each element is met.

YouTube is an interactive computer service, the complaint says, citing Gonzalez v. Google (9th Cir. 2021) and Lewis v. Google (N.D. Cal. 2020). The videos are provided by third parties, it adds, and the feed definition itself applies only to media "generated or shared by users." The contested element is the middle one: whether AB 1709 treats YouTube as a publisher.

Google's answer rests heavily on a 2026 Ninth Circuit decision, Doe 1 v. Meta Platforms, 174 F.4th 1159. The complaint returns to it repeatedly: "[M]atching users with content is publishing conduct, even when the user has not requested the content." On that reading, the personalization ban "imposes liability when YouTube chooses which third-party videos each user sees," the complaint argues, because liability "turns on what content it chooses for its users, and how it chooses that content." It also cites Force v. Facebook (2d Cir. 2019) for the proposition that websites act as publishers when they "present users with targeted content," even through algorithms.

Courts have not all read Section 230 that broadly. In August 2024 the Third Circuit held in Anderson v. TikTok that Section 230 does not shield a platform for its own conduct beyond hosting third-party content, and in April 2026 Massachusetts' highest court ruled that Section 230 did not bar the state's claims against Meta over design features that included autoplay and infinite scroll. Neither decision binds a federal court in California. Neither is mentioned in Google's complaint.

Autoplay as an act of publication

The autoplay ban gets its own line of reasoning. On YouTube, autoplay starts another user-uploaded video without a separate request. "Under the Act, YouTube must withhold the next video until the user affirmatively requests that it be played," the complaint says.

Google argues this regulates "the very act of publication itself," because "a video is published each and every time it is sent to a user," a point it supports with Barnes. Even if the ban were read only as a rule about how to publish, rather than whether to, the complaint says it would still fail: "Since the advent of the written word, publishers have had to decide not only what follows a given work, but whether it should follow immediately, later, or only upon request." It cites Turner Broadcasting v. FCC (1994) on the sequencing of video programming as editorial discretion, and a 2023 decision in the federal social media addiction litigation in the Northern District of California that held a defective-design claim about YouTube's "auto-play" was barred by Section 230.

Deleting accounts

The account-deletion requirement is treated as a third publishing decision. Forcing YouTube to close an account "it would prefer to keep," the complaint argues, compels it to "exercise editorial discretion" by "terminat[ing] an account." The authority it cites for that phrasing, Shared Partnership v. Meta Platforms, was decided by the Ninth Circuit on October 30, 2026 - five days before the complaint was filed.

The First Amendment argument

The constitutional count builds on Moody v. NetChoice (2024), in which the Supreme Court said that "[d]eciding on the third-party speech that will be included in or excluded from a compilation - and then organizing and presenting the included items - is expressive activity of its own." Google notes that Moody described YouTube's feeds directly, observing that the "selection and ranking" of videos "is most often based on a user's expressed interests and past activities."

The complaint makes two First Amendment claims. AB 1709 restricts YouTube's own expression in compiling third-party videos, it argues, and it burdens the right of under-16 users to receive that compilation. It quotes Brown v. Entertainment Merchants Association (2011): "[m]inors are entitled to a significant measure of First Amendment protection."

Human judgment inside the algorithm

This part of the complaint responds directly to the August ruling. In the SB 976 case, the district court held that YouTube had failed to show "that [its] personalized feeds convey an expressive message or reflect human editorial judgment," reasoning that because YouTube "rel[ies] on predictive algorithms that incorporate users' past watch history and other data," the resulting feeds did not reflect YouTube's expression.

The new complaint answers with detail about how people shape YouTube's recommender system, which the company calls the Recommender System and uses for its homepage, search results and autoplay. It quotes the Ninth Circuit's own 2025 language from NetChoice v. Bonta: "the more an algorithm implements human editorial directions, the more likely it is to be expressive," and "[a]n algorithm that promotes a platform's own message to users is likely to be protected speech."

The example it chooses is novelty. YouTube says it has decided that showing users unfamiliar content generates more long-term value than showing them what they already like, and that implementing this requires employees to make "subjective judgments like whether physics and math are sufficiently different so that they count as diverse or should be grouped together as STEM topics." Employees also "handpick training examples that teach the algorithm what qualifies as diverse," the complaint says. It ties this directly to the law: "YouTube cannot determine what will be novel and diverse to a user without considering what she has viewed before."

Content-based or not?

Google argues the ban is content-based and must survive strict scrutiny, because it targets speech "likely to be interesting, relevant, and valuable to an individual user." It also cites TikTok v. Garland (2025) for the rule that a law is content-based where speakers can "avoid or mitigate" its effects by altering their speech. To comply, YouTube says, it would have to change the compilation of videos in its feeds.

The complaint goes further, characterizing the law's purpose as suppressing speech. Citing the Act's legislative findings, it says the state wants to cut the time minors spend on social media "by forcing platforms to organize or display that speech in a less interesting way," which it likens to "a law requiring libraries to mix books up on shelves so minors cannot find ones that interest them and spend less time reading." That is Google's description of the findings, not the legislature's.

A separate line of argument concerns parental consent. SB 976, the 2024 law, lets users under 16 receive personalized feeds if their parents consent. AB 1709 removes that option. According to the complaint, in defending SB 976 the state "emphasized that parents retained the ability to override the State's prohibition," presenting the consent mechanism as its way of restricting as little speech as possible. "The State cannot now claim that a complete ban on personalized feeds that strips YouTube of editorial control and overrides parents' assessment of what is best for their children is necessary to achieve that same interest," the complaint says.

It also points to a brief the state filed in the Ninth Circuit on October 27, 2026, which, according to Google, "affirmatively endorsed the less restrictive alternative of providing minor users with an option for non-personalized feeds." "A complete ban is the antithesis of narrow tailoring," the complaint says.

On fit, it calls the law "vastly overinclusive," since recommending "videos about mechanical engineering to a teen interested in robotics" is caught along with everything else. And it calls the law underinclusive for leaving alone other platforms that deliver video through personalized feeds, naming Netflix and Hulu, without providing any "persuasive reason why."

The evidence question

Under strict scrutiny, the complaint says, the state must "show a direct causal link" between personalized feeds and "harm to minors." Google claims the state cannot. "In defending SB 976, the State produced over 500 pages of expert materials and yet was unable to identify a single peer-reviewed or published study establishing that personalization causes or is correlated with negative health outcomes in minors," the complaint says. It adds that "the State's only study that specifically considered YouTube found no link between personalization on YouTube and harm to minors."

It cites a declaration filed on February 20, 2026 by Dr. Jenny S. Radesky, one of the state's experts in the SB 976 case, for the claim that only a small number of minors suffer from "[s]ocial media addiction." These are Google's characterizations of the state's evidence; the declaration and the expert materials themselves are not part of this complaint.

How YouTube describes its recommendations

Much of the 28 pages is a description of YouTube itself. According to the complaint, YouTube holds more than 20 billion videos, adds more than 20 million every day, and is available in over 100 countries and 80 languages. "Beginning in March 2026," it says, YouTube "has accounted for a higher percentage of television use among U.S. viewers than any other media distributor in the United States," and it is "the most frequently used service for listening to podcasts in the United States." Nielsen's figures point the same way: YouTube took a record 15.8% of US TV time in Nielsen's August 2026 Gauge.

Most of what it hosts is user-generated content: "The vast majority of content on YouTube is created, edited, uploaded, and controlled by third parties." The company says it does not rank videos by recency, by generic popularity or by a social graph. Chronological feeds would bury older high-quality videos, it argues, and "bad actors may be able to take advantage of chronological ordering to spread potentially harmful content to more users." A feed sorted by popularity would push content that is "trending" across all ages, which may be unsuitable for younger users.

Inputs to the system, as the complaint lists them, include a user's age, language, location, watch and search history, videos enjoyed by similar users, survey responses, likes, dislikes, the "not interested" button and channel subscriptions. On top of that sit YouTube's own standards: it deprioritizes clickbait and low-rated videos and raises videos "from trustworthy sources or recent videos on newsworthy topics." The work involves "hundreds of employees who set metrics for success and evaluate the system against those metrics," test versions and adjust the system.

One detail has a direct bearing on the "search instead" argument the August ruling relied on. According to the complaint, "YouTube designed its Search function so that a user who enters a search query receives results that, in addition to being generally high-quality, are personally relevant to that user." The complaint does not say whether YouTube believes AB 1709 reaches personalized search results.

What is already different for minors

The complaint sets out what YouTube says it already does for younger users:

  • Content in certain categories that does not break the Community Guidelines, such as nudity and sexually suggestive material, violent or graphic content and vulgar language, is age-gated and not recommended to users under 18.
  • For teens aged 13 to 17, YouTube avoids repeatedly recommending content that is harmless once but could be problematic in volume, "such as content that displays social aggression or idealizes certain body types."
  • For under-13s, it promotes content identified with outside experts, such as grade-appropriate educational videos, and deprioritizes content it deems "low quality" for children, giving "heavily promotional or commercial content" as the first example.
  • Autoplay "is off by default for every user under 18," and parents can switch it off entirely for supervised children under 13.
  • Supervised teens are placed in a "Teen" content setting for ages 13 to 15 and "Older teen" for 16 and 17. Parents of supervised kids and teens can set a daily YouTube Shorts limit between 0 minutes and 2 hours. Google announced the zero-minute Shorts timer in February 2026.

Removing personalization, the complaint argues, "would increase the risk that young users are exposed to content that is not appropriate for them." That claim runs directly against the premise of the law.

A second run at California after August

AB 1709 is California's second law on minors' feeds. SB 976, enacted in September 2024, bars personalized feeds for minors without parental consent. NetChoice, a trade group of which Google is a member, challenged it in November 2024. The district court denied a preliminary injunction on the feed provisions on the ground that NetChoice lacked associational standing, and in September 2025 the Ninth Circuit affirmed, saying "the unique design of each platform and its algorithm affects whether the algorithm at issue is expressive."

That ruling invited individual companies to sue. Google, YouTube and Meta filed separate lawsuits on November 13, 2025, and TikTok brought its own challenge the same day. Across "several rounds of briefing, numerous fact and expert declarations, and a lengthy hearing," YouTube says it provided extensive evidence that its feeds are expressive.

It lost. On August 5, 2026, Judge Edward J. Davila denied preliminary injunctions to Meta, TikTok and Google. According to the new complaint, the court acknowledged that YouTube's feeds seek to convey that "the user will likely find the displayed content interesting and enriching," but held that the reliance on predictive algorithms meant the feeds did not reflect YouTube's expression, and that "minor users remain free to search for specific content."

YouTube appealed on August 11 and asked the district court for an injunction pending appeal, which was denied on August 21. On August 24 it asked the Ninth Circuit for the same relief. On September 17 a motions panel referred that request to the merits panel and put the appeal on the next available oral argument calendar. The complaint says the appeal is ongoing.

What is different this time? Section 230. The August ruling, as described in the complaint, turned on whether the feeds were expressive. A preemption claim does not depend on that question. It asks whether the state is treating YouTube as a publisher of third-party content, and Doe 1, decided in 2026, gives Google Ninth Circuit language to work with. The autoplay ban is challenged only on that ground.

Meta filed the same day

Meta Platforms filed its own 49-page complaint against Bonta on November 4, case 3:26-cv-13201 in the same court, challenging AB 1709 as applied to Facebook, Instagram and Threads. Meta, represented by Covington & Burling, argues that the personalized-feed ban and the autoplay ban violate the First Amendment, and that both bans and the account-deletion requirement are preempted by Section 230; it also cites Doe 1 and Shared Partnership. One difference stands out. Meta challenges the autoplay ban on First Amendment grounds as well, while Google confines its constitutional count to personalized feeds.

Why this matters for the marketing industry

For advertisers, the case is not about ad targeting. The complaint does not discuss advertising, and the law as Google describes it reaches feeds of media "generated or shared by users." Google already limits ads shown to minors: since July 2025 it has been using machine learning to estimate users' ages and switching off ad personalization for those likely under 18.

The commercial exposure sits elsewhere. YouTube says personalized recommendations "allow the millions of content creators on YouTube to gain exposure, share their ideas, and connect with their target audiences." If AB 1709 applies to YouTube from January 1, 2027, reach among Californians under 16 would depend on non-personalized feeds and on search rather than on recommendations. That affects creators, toy and gaming brands, educational publishers and anyone running branded channels for young audiences. YouTube's statement that it already deprioritizes "heavily promotional or commercial content" for under-13s shows that commercial content is already ranked down for its youngest users.

Then there is the account-deletion rule. Deleting under-16 accounts presumes platforms know who is under 16. AB 1709 leans on California's Digital Age Assurance Act, AB 1043, which also takes effect on January 1, 2027. That makes age assurance part of the compliance picture for platforms with young users in California, however this case ends.

Alphabet does not break out what is at stake. YouTube's ad revenue rose 9% to $11.2 billion in the third quarter of 2026, its slowest growth of the year. The complaint gives no figure for the share of users or revenue coming from Californians under 16, and Alphabet does not report one.

California is not acting alone. Australia's under-16 social media ban took effect on December 10, 2025 and covers YouTube. The UK government named YouTube among the platforms in its planned under-16 ban, due in spring 2027. Turkey's regulation, in force since November 1, 2026, bars social media accounts for under-15s and restricts interfaces that serve visual content one item after another, including "automatically." In the EU, the European Commission's July 2026 preliminary finding against Meta named autoplay and personalised recommender systems among the design features it objected to.

Courts are moving too. A Los Angeles jury in March 2026 found Meta and YouTube negligent in a social media addiction case, awarding $6 million, of which YouTube's share was $1.8 million. That was a damages case; Google v. Bonta is about whether a state can dictate feed design in advance. A ruling for Google on Section 230 could reach beyond YouTube, and Meta is already making the same argument. A ruling against it would leave California free to regulate how recommendations work for under-16s, not only what they show.

Timeline

Summary

Who: Google LLC and YouTube, LLC, represented by Cooley LLP with Elizabeth B. Prelogar signing, sued California Attorney General Rob Bonta in his official capacity. Meta Platforms filed a separate challenge to the same law on the same day.

What: A 28-page complaint for declaratory and injunctive relief claiming that Section 230 preempts AB 1709's bans on personalized feeds and autoplay for users under 16 and its account-deletion rule, and that the feed provisions violate the First Amendment as applied to YouTube. It seeks a preliminary and permanent injunction against enforcement against YouTube.

When: Filed on November 4, 2026. AB 1709 was signed on September 10, 2026 and takes effect on January 1, 2027. YouTube's appeal against the August 5, 2026 ruling on SB 976 is still pending in the Ninth Circuit.

Where: The US District Court for the Northern District of California, San Jose Division, case 5:26-cv-13204-SVK. The law applies to users in California.

Why: AB 1709 removes the parental-consent route that SB 976 allowed and adds an autoplay ban and penalties of up to $50,000 per affected minor. After losing the First Amendment argument at the preliminary stage in August, Google is adding a federal preemption claim. For creators, brands and publishers reaching young audiences on YouTube, the outcome decides whether recommendation-driven reach among Californians under 16 survives into 2027.