Platform Suppression
A series of landmark rulings signals that Section 230 immunity was never meant to cover deliberate product design choices — and Silicon Valley is starting to feel it.
NewsOnScale Staff
September 12, 2026
Something significant is happening in American courtrooms, and it is moving faster than most technology policy observers expected. Judges and juries are increasingly drawing a distinction that platforms have spent years trying to obscure: the difference between what users post and what platforms deliberately engineer. The legal theory underlying recent verdicts against Meta and Google is not about speech. It is about design.
That distinction matters enormously, and not just for lawyers.
## What Section 230 Actually Says — and What It Doesn't
Section 230 of the Communications Decency Act, passed in 1996, provides internet platforms with immunity from liability for content created by third-party users. The intent was straightforward: don't hold the phone company responsible for what people say on calls. In the early internet era, that logic helped an industry grow without being crushed by litigation before it could find its footing.
But the law was never written to immunize a company for choices it makes itself. When a platform decides to deploy an autoplay video queue, an engagement-maximizing recommendation algorithm, or a notification system calibrated to exploit psychological vulnerability, those are not user decisions. They are product decisions. Increasingly, courts are treating them that way.
The Massachusetts Supreme Judicial Court's recent ruling is among the clearest articulations of this principle yet from a state high court. The decision holds that plaintiffs can pursue claims rooted in platform design even when Section 230 might block claims about specific pieces of user-generated content. The legal surface area for accountability just expanded considerably.
## Why This Wave Is Different
Platform design litigation is not new. What is new is that it is winning. Historic verdicts against Meta and Google — the details of which are still reverberating through industry legal teams — signal that juries are willing to hold companies responsible when the architecture of their products can be shown to cause foreseeable harm.
This matters for several reasons beyond the immediate dollar figures involved. First, it creates precedent. Appellate courts consolidating these rulings will produce binding legal standards that reshape what platform engineers are allowed to build without consequence. Second, it creates discovery. Litigation forces internal documents into the record. The gap between what these companies say publicly about their products and what their own researchers documented privately has, in case after case, proven to be substantial.
Third — and this is the piece that gets underreported — it changes the incentive structure. If engagement-maximizing design features become a source of significant legal liability, the calculus for building them shifts. Regulation often moves slowly. Litigation, especially when plaintiffs start winning, moves at the speed of the next quarterly earnings call.
## The AI Parallel Is Already Here
The timing of this litigation wave is not incidental. As courts work through questions of platform design liability for social media, a parallel debate is opening about AI systems. The same core question applies: when a company makes deliberate architectural choices about how an automated system behaves, and harm results, can it claim immunity because a third party's data or input was somewhere in the chain?
The emerging legal consensus, at least in the social media context, suggests the answer is no. Scholars and civil society organizations are now explicitly arguing that AI liability frameworks should be built on this foundation from the start, rather than allowing a new generation of platforms to spend another thirty years litigating the same evasions under new branding.
## What Accountability Actually Requires
None of this means Section 230 is or should be dead. The law still serves legitimate purposes for smaller platforms and open communication infrastructure that genuinely cannot screen every piece of user content. The question is whether the largest, most profitable technology companies — organizations that employ thousands of engineers to optimize user behavior at scale — ever deserved to hide behind it for the consequences of their own product choices.
The courts are beginning to answer that question. The answer, increasingly, is no.