Social media giants fail to block California prohibition of addictive feeds for minors

The companies failed to persuade a judge that their reliance on algorithms amounted to “expressive” activity meriting First Amendment protection. The logos of Facebook, YouTube, TikTok and Snapchat on mobile devices from 2017-2022. (AP Photo, File) (CN) — A federal judge rejected on Thursday a bid by Meta, Google and TikTok to block provisions of California’s Protecting Our Kids from Social Media Addiction Act that prohibits them from sending minors personalized — or addictive — feeds. Senior U.S. District Judge Edward Davila in San Jose, California, denied the social media behemoths’ request for a preliminary injunction against enforcement of the statute’s provisions, which bar them from sending minor users personalized feeds based on their prior interactions with their platforms. The Barack Obama appointee said the three companies weren’t likely to succeed with their First Amendment challenge to the state law because the personalized feeds aren’t “expressive,” but the product of predictive algorithms that incorporate a user’s past watch history and other data to suggest content that the algorithms anticipate will be engaging or interesting to this user. “Plaintiffs are not making any decisions about what content will be ‘interesting,’ because they are merely relying on predictive modeling to assess what users’ characteristics and history on the platform suggest will keep these same users engaged,” Davila said. “This decision is not an ’expressive’ message; it is merely a mirror that reflects back to users their own perceived interests.” The judge noted his decision pertains only to the companies’ request for a preliminary injunction and that, once they have had a chance to present him with all the facts in support of their arguments, he could conclude the law’s provisions are indeed unconstitutional. Representatives of the social media platforms had no immediate comment on the ruling. The three companies argued at a June hearing that restricting recommended content — both the human decisions about what to put in their social media feeds and the algorithms that determine what users see as they scroll — regulates protected expression and imposes unconstitutional burdens on First Amendment-protected actions. YouTube attorney Elizabeth Prelogar, a former solicitor general under the Biden administration, said the algorithms and decisions involved in recommending videos or organizing a 13-year-old’s feed to reflect her most viewed content based on engagement are part and parcel to content moderation and cannot be parsed. “Humans have a tight, close connection and link to how the outputs match up with the inputs and the system is functioning to match the editorial goals,” she said. Prelogar said personalized feeds are especially important for minors, as they recommend new, interesting and age-appropriate content and are how — through user signals, subject matter and viewership — online platforms will know what to recommend based on whether a user is a teenager or a child under 13 years old. But Davila indicated at the hearing that he was concerned about artificial intelligence’s role in creating the recommendations and ultimately curating feeds that may be inappropriate or harmful to minors. California Deputy Attorney General Shiwon Choe told him that the platforms’ algorithms “fill in the blank” without human direction and without “actually understanding the content or messages” that are selected, aiming to increase engagement — and more ad viewership and, therefore, more money for the platforms. California contends the unknowability of algorithms’ recommendations harms children. Choe cited the example of a minor girl going unconscious after choking herself because of a “blackout challenge,” a social media trend of filming a participant experiencing asphyxiation and sharing the content, based on recommended videos in her feed. The onlin