The numbers come first. Meta says it faces theoretical penalties of $1.4 trillion as four states — California, Colorado, Kentucky and New Jersey — pursue claims that the company deliberately designed Instagram and Facebook to keep young users engaged past the point of healthy use. The trial is unfolding in Oakland, and the features under scrutiny are not edge cases: infinite scroll, autoplay, push notifications and recommendation algorithms sit at the core of how both platforms function.
Critics have reached for the Big Tobacco comparison, and Meta's own legal exposure invites it. But the analogy has limits that matter for anyone thinking about what a court-ordered remedy would actually look like.
Larry Magid, an online-safety advocate who has advised Meta on safety issues since 2005 and currently serves on the company's Safety Advisory Council, Youth Advisory Council and a safety advisory group for Meta Reality Labs, spent years preferring a different frame. He told Fortune that social media, unlike cigarettes, carries genuine benefits: young people use it to maintain friendships, find communities, organize around causes and access information. His preferred metaphor was chocolate — pleasurable, potentially excessive, but not inherently harmful.
The algorithms changed his thinking. 'The algorithms change the whole nature of it, and that's where my chocolate metaphor went awry,' Magid told Fortune. A chocolate bar does not refill itself. Infinite scroll and autoplay do something closer to that — Magid likened the dynamic to a grocery store automatically delivering more chocolate bars whenever it notices the supply running low.
The dilemma for Meta is that stripping out the recommendation engine is not a clean fix. Magid tested it himself, switching his own Facebook feed to chronological order. He found it boring and returned to the algorithmic feed. 'I actually went back to the algorithm because I actually found it was benefiting me in some ways,' he said.
His proposed middle ground: make recommendations less aggressive and weight them more heavily toward a user's 'social graph' — the friends, communities and interests a person has actively chosen — rather than toward content optimized purely for time-on-platform. That would represent, in his framing, a partial return to what social media originally was before platforms shifted from friend-to-friend interaction toward algorithmically curated feeds driven by influencers and strangers.
The gap between what an adult sees and what a teenager sees on the same platform is also part of the record. Magid said his own feeds surface aviation content, news and occasional political debate. But he said minors he speaks with describe a markedly different experience. 'I am told by minors, by high school kids and young adults, that they have a very different experience than I do,' Magid said. 'They are seeing misogyny, they're seeing homophobia, they're seeing racism.' The same recommendation engine that learns Magid likes airplanes can learn what captures a teenager's attention.
CEO Times take: The trial in Oakland is, at its core, a question about product liability in the digital age — and the answer will carry real costs regardless of outcome. If Meta loses, the $1.4 trillion exposure figure alone is a systemic risk to shareholders and to the broader ad-supported internet economy. If regulators or courts mandate feature restrictions, the cure may degrade a product that hundreds of millions of adults use voluntarily and value. The better path — one that free enterprise can actually execute — is transparent parental controls, age-appropriate algorithm defaults and clear liability rules, not a courtroom redesign of a platform's core architecture. Capital rewards clear rules. What it cannot price is an open-ended judicial mandate to make a product less useful.
