Netchoice, LLC v. Bonta
Issue: Whether California's Age-Appropriate Design Code Act (CAADCA), Cal. Civ. Code §§ 1798.99.28–1798.99.40, facially violates the First Amendment through its coverage definition, age estimation requirement, data use restrictions, and dark patterns prohibition, as evaluated under the *Moody v. NetChoice* standard for facial challenges.
On remand from a prior Ninth Circuit decision (*NetChoice I*) that had vacated most of the initial preliminary injunction for failure to apply the correct facial-challenge standard under *Moody v. NetChoice*, 603 U.S. 707 (2024), the district court again enjoined the CAADCA in its entirety, concluding that the statute's coverage definition was content-based and triggered strict scrutiny throughout. The Ninth Circuit, reviewing for abuse of discretion, vacated the injunction as to the coverage definition and age estimation requirement because NetChoice failed to develop a sufficient record to satisfy *Moody*'s two-step facial-challenge framework—specifically, NetChoice did not adequately catalog the full scope of the law's applications to allow the court to assess what share of those applications were unconstitutional. The panel affirmed the injunction of the data use restrictions and dark patterns prohibition solely on void-for-vagueness grounds, finding those provisions fail to delineate proscribed conduct with sufficient clarity, and vacated the severability ruling as to the statute's remaining provisions because the record did not support a likelihood that valid provisions were inseverable from the enjoined notice-and-cure provision.
The decision reinforces that First Amendment facial challengers—including sophisticated litigants like NetChoice—bear a demanding burden under *Moody* to build a record mapping a law's full set of applications before courts can measure unconstitutional uses against the statute's legitimate sweep, effectively raising the evidentiary threshold for pre-enforcement facial injunctions against online child-safety laws. The ruling also signals that states retain meaningful room to enact children's digital privacy legislation, at least where challengers cannot demonstrate facial invalidity across a substantial majority of the law's applications.
Issue: NetChoice v. Bonta* asks whether a facial First Amendment challenge can sustain a wholesale injunction against California's Age-Appropriate Design Code Act when the statute's coverage definition reaches businesses through both content-based and purely demographic indicators — meaning some covered services may have no expressive dimension at all. The case also asks whether terms like "best interests of children" and "materially detrimental," borrowed from individualized family-law proceedings, are unconstitutionally vague when applied as prospective, industry-wide compliance standards. The answers turn on how courts measure the proportion of unconstitutional applications against all applications under the demanding framework the Supreme Court established in *Moody v. NetChoice* (2024).
This is a published, precedential Ninth Circuit opinion resolving California's appeal from a district court order that, on remand from an earlier Ninth Circuit decision, again enjoined the CAADCA in its entirety. The three-judge panel unanimously affirmed in part and vacated in part. It affirmed the injunction of the data-use restrictions and dark-patterns prohibition on vagueness grounds, holding that welfare-standard terms like "best interests of children" are too indeterminate to provide fair notice for class-wide regulatory compliance and that the statute's scienter element does not cure that defect. The court vacated the whole-statute injunction, however, finding that NetChoice had failed to catalog the Act's full range of applications — by focusing exclusively on content-publishing platforms, it ignored non-content coverage indicators that could reach ride-sharing or financial services, effectively converting a facial challenge into an as-applied one. The panel also vacated and remanded the injunctions concerning the age-estimation requirement and the severability of the valid remainder, finding both questions unresolved on the existing record.
This ruling significantly raises the evidentiary bar for industry coalitions seeking to block child online-safety laws through facial First Amendment challenges: plaintiffs must now map a statute's *entire* universe of applications — including non-expressive ones — before a court can find unconstitutional applications substantial enough to justify a wholesale injunction. The vagueness holding breaks new ground by applying *FCC v. Fox*'s void-for-vagueness standard to child-welfare design mandates, establishing that family-law welfare terms cannot be transplanted into prospective regulatory compliance obligations without adequate definitional grounding. Significant questions remain open on remand, including whether the age-estimation requirement implicates First Amendment-protected speech and whether the CAADCA's valid provisions are severable — determinations that will shape California's enforcement posture and may influence how courts in other circuits assess analogous age-gating laws.