NetChoice v. Bonta (CA Addictive Feeds Law)

Questions Presented

Is it a company’s protected expression under the First Amendment to present a user with whatever content the company’s algorithm thinks will maximize the amount of time the user spends on the company’s website?

Does blocking a social media website from sending push notifications to minors during school hours and late at night constitute a First Amendment violation?

Does a regulation requiring a social media company to publish statistics about the number of minor users on its website and their aggregate settings choices violate the First Amendment?

Does a law requiring an online service provider to estimate a user’s age before allowing them to use certain website features violate the First Amendment?

Background

Legislatures are increasingly focused on protecting kids from online harms. Recently, some legislatures have passed laws regulating how companies design features on their platforms. These laws tend to target harmful, content-agnostic features on social media platforms, so they raise fewer First Amendment concerns than laws seeking to regulate the content available to users who visit a platform.

Recently, the California and New York legislatures have passed laws regulating addictive design features on social media platforms that can interfere with kids’ social lives, sleep, studies, and more. Examples of such features include engagement-maximizing content recommendation algorithms and vague, frequent push notifications that repeatedly nudge minors to return to the platform.

Tech companies have been fighting back against these laws, and virtually all laws regulating their activities, through extensive lobbying and lawsuits. Many of the tech companies’ lawsuits take a specific form: the pre-enforcement facial challenge. These kinds of challenges seek to prevent a law from being enforced against anybody by claiming that, no matter how the law is enforced, it is guaranteed to violate speech rights at scale.

Facial challenges are supposed to be unusual and relatively rare because they bend the rules of normal litigation: instead of a party arguing that a law is actually being enforced against them in an unconstitutional way, they argue that the law will likely be unconstitutional when enforced against anybody. In Moody v. NetChoice and NetChoice v. Paxton,a recent pair of cases involving facial challenges, the Supreme Court remarked that “For a host of good reasons, courts usually handle constitutional claims case by case, not en masse. Claims of facial invalidity often rest on speculation about the law’s coverage and its future enforcement. And facial challenges threaten to short circuit the democratic process by preventing duly enacted laws from being implemented in constitutional ways.” For this reason, parties bringing facial challenges are supposed to construct detailed, comprehensive records that allow courts to determine whether, among a law’s hypothetical applications, the unconstitutional ones substantially outweigh the constitutional ones.

In contrast to the requirements set out in Moody and Paxton, tech companies’ facial First Amendment challenges often rely on oversimplifications, sloppy analogies, and barebones records in an attempt to garner overly broad First Amendment protections for their business practices. For example, companies often attempt to claim, without supporting evidence, that organizing content via black-box algorithms is no different from a newspaper’s editorial judgment in placing news stories, that any law requiring a website to treat children and adults differently will chill adults’ right to access speech, and that any transparency requirements around their activities requires them to opine on controversial topics.

The Case

In this case, NetChoice—a tech industry trade association—is suing to stop the enforcement of California’s “Protecting Our Kids from Social Media Addiction Act” (SB 976). This law seeks to protect kids from a specific harm: social media addiction that interferes with sleep, learning, in-person socializing, etc. To do this, the law prevents websites from using certain addictive design features on minors without parental consent. The regulated design features include engagement-maximizing algorithms and nighttime push notifications. The law also compels social media companies to disclose statistics such as the number of minors using the platforms and how many minors’ parents have consented to addictive design features. For the next two years, companies only have to avoid using such features on users they actually know are minors. After that, the California Attorney General is tasked with issuing regulations on reasonable age estimation methods that companies must use to estimate which users are children.

NetChoice brought a facial and as-applied First Amendment challenges arguing that the whole law violates the First Amendment. For example, it argues that, by regulating engagement-maximizing algorithms that in part create social media feeds, the law interferes with the companies’ protected editorial judgment. It also argues that any law ordering social media companies to distinguish between minors and adults violates the First Amendment.

EPIC’s Brief

EPIC submitted a brief in support of Bonta’s opposition to NetChoice’s motion for a preliminary injunction, which would have prevented enforcement of the law while the lawsuit made its way through the courts.

EPIC’s brief explained that NetChoice’s motion should be denied because NetChoice has not carried its burden under Moody and Paxton to develop a detailed factual record or to make specific legal arguments to support any of its First Amendment claims. NetChoice did not explain how its members’ feeds work, why those feeds should be considered expressive, how the challenged law would inhibit that expression, or how the age assurance that the Attorney General selects in two years will burden expression.

EPIC’s brief also explained why the Court should view NetChoice’s claim that the law regulates expression with skepticism. In Moody and Paxton, the Supreme Court said that social media feeds themselves are expressive products, but not every activity taken to create those feeds is itself expressive. For example, it is likely expressive when companies remove or downrank content that violates their content and community guidelines, such as racist posts. The Court found this similar to what newspaper editors do and likely expressive. But it is not likely expressive when companies surveil users’ every move to construct profiles of them, feed that information into algorithms, and tell the algorithms to show users whatever content similar users spent the most time engaging with. The Court explicitly called this out as a practice that might not be expressive in Moody/Paxton, and it is is what SB 976 regulates, so NetChoice’s citations to Moody are inappropriate.

EPIC’s brief also explained why the Court should decline to rule that the law’s age assurance provisions are unconstitutional. The Supreme Court has repeatedly held that an age assurance requirement’s constitutionality depends on factual specifics. Since SB 976 does not require age assurance for two years and leaves it to the Attorney General to declare which age assurance methods are reasonable and how they should be regulated, the court cannot yet rule on whether it is constitutional.

Litigation Documents