I recently came across a Reuters[1] article titled “Meta must face youth addiction lawsuit by Massachusetts, court rules”, reporting on a decision of the Massachusetts Supreme Judicial Court that, in my view, is far more significant than the headline suggests. In Commonwealth v. Meta Platforms, Inc. (2026) (SJC-13747)[2], the Court refused to dismiss a lawsuit alleging that Instagram was deliberately designed to induce compulsive use in children, not through the content it hosts, but through the architecture of the platform itself. The Commonwealth’s claim is direct and, if ultimately proven, potentially transformative, alleging that features such as infinite scroll, push notifications, and variable reward mechanisms were engineered to exploit the neurological vulnerabilities of young users, prolong screen time, and drive advertising revenue.
What matters at this stage is not that liability has been established. It hasn’t. But that the Court drew a clear and deliberate line between liability for third-party content, which remains broadly protected, and liability for a platform’s own design choices and representations. In doing so, the Court allowed the case to proceed on the basis that the alleged harm flows from Meta’s conduct as a product designer, not merely as a publisher. That distinction, still developing and far from settled, provides a useful starting point for examining how courts are beginning to grapple with social media, addiction, and the outer limits of legal responsibility.
That distinction reflects a broader and growing instinct, particularly in the United States, to treat social media platforms not as neutral conduits, but as products deliberately engineered to capture, hold, and shape human attention. As a grandfather of seven curious children, the issue of phone addiction is not abstract to me, and it is something I am watching closely. That instinct is now working its way into courtrooms. But if you strip away the headlines, the legal reality is more restrained, and far more unsettled, than the commentary suggests. What follows is a careful walk through the actual case law that exists today, what American courts have and have not done with it, and what, if anything, it might mean north of the border.
The U.S. Litigation Landscape: Ambition Outpacing Results
The idea that social media companies can be held liable for “addictive design” has been tested, but not yet proven in the way many assume. A frequently cited decision I see is Lemmon v. Snap Inc., 995 F.3d 1085 (9th Cir. 2021)[3], arising from a tragic set of facts in which three teenagers were killed in a high-speed crash after using Snapchat’s “Speed Filter” to record how fast they were travelling. The plaintiffs argued that Snap had designed a product that foreseeably encouraged dangerous behaviour, an argument that now sits at the heart of many “addictive design” cases involving young users. Unlike many other cases in this area, the Ninth Circuit did not shut the claim down at the outset. Instead, it held that section 230 of the Communications Decency Act of 1996, 47 U.S.C. § 230[4] (providing immunity to online platforms from liability for third-party content) did not apply, because the claim did not seek to treat Snap as a publisher of third-party content, but rather as a product designer. The Court emphasized that the alleged duty to design a reasonably safe product was independent of Snap’s role in hosting or transmitting user content, and that the harm flowed from the platform’s design features themselves, not from any specific content created by users. In doing so, the Court allowed the claim to proceed, marking a clear judicial recognition, at least at the pleadings stage, that design-based claims against social media platforms may fall outside the traditional shield of publisher immunity.
That same theme appears in Force v. Facebook, Inc., 934 F.3d 53 (2d Cir. 2019), where plaintiffs who were victims and families of victims of Hamas terrorist attacks, alleged that Facebook’s platform and algorithms facilitated those attacks by allowing Hamas to disseminate propaganda and by recommending content and connections to users. The Second Circuit accepted, at the pleadings stage, that Hamas used Facebook to post and amplify messages encouraging violence, and that attackers may have been exposed to that content through Facebook’s systems. Even so, the Court held that these functions of hosting user content, organizing it, and directing it to interested users through algorithms, remain quintessential publishing activities. In the Court’s view, algorithmic amplification did not transform Facebook into a creator or developer of the underlying content, but merely reflected the automated exercise of traditional editorial functions. As a result, the claims necessarily treated Facebook as the publisher of third-party information, and were therefore barred by s. 230 of the Communications Decency Act.
The issue reached the U.S. Supreme Court in Gonzalez v. Google LLC, 598 U.S. ___ (2023)[5], arising out of the 2015 ISIS attacks in Paris that killed 130 people, including a U.S. citizen. The plaintiffs alleged that Google, through YouTube, played a role in radicalization and recruitment by hosting ISIS content and, more importantly, by recommending that content to users through its algorithms. The theory was a familiar one that recommendation systems do more than passively display content, they actively connect it to susceptible audiences. The Ninth Circuit largely held that such claims were barred by section 230, with limited exceptions that ultimately failed on the merits. When the case reached the Supreme Court, however, the Court declined to resolve the broader question. Instead, it focused on a threshold issue of whether the plaintiffs had even stated a viable claim for aiding and abetting terrorism under the Anti-Terrorism Act, 18 U.S.C. § 2333[6] (providing a civil cause of action for injuries arising from acts of international terrorism). Relying on its companion decision in Twitter, Inc. v. Taamneh, 598 U.S. ___ (2023)[7], the Court concluded that the allegations, taken at their highest, did not plausibly establish that Google had knowingly provided substantial assistance to ISIS. In light of that conclusion, the Court sidestepped the section 230 question entirely, vacating the judgment and remanding the case without addressing the scope of platform immunity.
As far as I can see, that is the current state of play south of the border. The theory is advancing, but the law has not caught up.
The Emerging “Addictive Design” Cases
There is, however, a second wave of litigation, particularly involving minors, that is attempting to bypass section 230 altogether by framing social media platforms as defective products.
These cases have been consolidated in the U.S. federal system as the In re Social Media Adolescent Addiction / Personal Injury Products Liability Litigation (MDL No. 3047, N.D. Cal.)[8] in the Northern District of California, before Judge Yvonne Gonzalez Rogers. What began as a series of individual lawsuits has now grown into hundreds of claims brought by children, parents, school districts, and state attorneys general, all advancing a common theory that the harm does not arise from isolated pieces of content, but from the design of the platforms themselves.
The plaintiffs allege that platforms such as Instagram, TikTok, Snapchat, and YouTube were not merely passive hosts of user expression, but were deliberately engineered to foster compulsive use among minors. According to the pleadings, these platforms deploy features such as infinite scroll, algorithmic recommendations, intermittent reward systems, and persistent notifications to exploit well-understood neurological and psychological vulnerabilities in young users. The result, it is said, is not simply increased engagement, but measurable harm, including rising rates of anxiety, depression, sleep disruption, eating disorders, and, in some cases, self-harm.
What is striking and legally significant to me in reading these cases as a Canadian negligence lawyer is how the claims are framed. The plaintiffs do not primarily challenge the content appearing on these platforms. Instead, they allege that the platforms themselves function as defective products, and are systems designed to maximize time-on-device at the expense of user well-being, without adequate warnings of the risks. The Judicial Panel on Multidistrict Litigation, in transferring and consolidating these cases, noted that the actions share common factual questions, including whether the defendants knew of these risks and failed to take reasonable steps to mitigate or disclose them.
That framing matters. It places the litigation squarely within the emerging fault line between traditional publisher immunity and product-based liability, a distinction that, as the earlier cases suggest, remains unsettled but increasingly difficult for courts to ignore.
What matters for present purposes is that there has not yet been a definitive trial judgment establishing liability on that theory. Motions to dismiss have been partially denied in some instances, allowing claims to proceed, but that is a long way from a finding of liability after a full evidentiary record.
That distinction is critical. Allowing a claim to proceed means only that it is arguable. It does not mean it will succeed.
The Legal Fault Line: Content vs. Design
The entire debate turns on a single legal fault line: whether the harm flows from content or from design. If the harm is characterized as arising from third-party content, section 230 protection in the United States is formidable. Courts have consistently refused to allow plaintiffs to repackage content-based claims as design defects. If, however, the harm can be framed as arising from the platform’s own conduct including its architecture, its reward systems and its algorithmic reinforcement loops, then the case begins to look more like a traditional products liability claim.
That is the path the current American litigation is trying to carve. But it is not an easy path. Courts are wary of collapsing the distinction between publisher and product. Once that line blurs, the entire legal foundation of the modern internet starts to shift.
What This Does and Does Not Mean for Canada
In Canada, the terrain is different, but not necessarily more favourable to plaintiffs.
Canada does not have a statutory equivalent to section 230 of the Communications Decency Act. Instead, the protection afforded to online intermediaries has developed incrementally through the common law, most notably in defamation. I wrote a recent Substack on the subject of defamation, “Loose Lips Sink Ships: Law of Defamation in Canada” [9].
In Crookes v. Newton, 2011 SCC 47[10], the Supreme Court of Canada held that the mere act of hyperlinking does not constitute publication, emphasizing that passive facilitation of content, without more, should not attract liability. That reasoning reflects a broader judicial instinct to avoid imposing responsibility on actors who do not meaningfully control or create the content at issue. But the decision also leaves open an important question. If liability turns on the distinction between passive facilitation and active involvement, where does algorithmic design fall? Unlike a simple hyperlink, modern platforms do not merely point users toward information. They shape, prioritize, and reinforce engagement in ways that are neither neutral nor entirely transparent. The absence of a statutory shield like section 230 does not answer that question, but it does leave Canadian courts with more doctrinal flexibility to confront it directly.
At first glance, that might suggest a more open field for claims against social media companies. In reality, Canadian law imposes its own constraints, most notably through the doctrines of negligence, proximity, and policy-based limits on duty of care.
The Supreme Court of Canada has repeatedly emphasized that recognizing a novel duty of care requires careful scrutiny, particularly where the claim engages broader policy considerations.[11] A court would have to ask whether it is fair, just, and reasonable to impose a duty on a platform to design its product in a way that avoids psychological harm to users, especially when user engagement is the very purpose of the product.
There are immediate obstacles.
First, foreseeability is not enough. The plaintiff would also have to establish proximity. That is difficult in a context where the relationship between platform and user is diffuse, standardized, and mediated through terms of service.
Second, causation would be a formidable hurdle. Unlike a defective physical product, where the chain of causation is often direct, social media harm is multi-factorial. Canadian courts would be reluctant to attribute complex psychological outcomes to platform design alone without clear, non-speculative evidence.
Third, there is the policy dimension. Imposing liability for “addictive design” would effectively require courts to regulate the architecture of digital platforms. That is a role courts have historically approached with caution[12].
I could not find, nor do I believe, that there is any current Canadian case law establishing liability against a social media company on an addictive design theory. That silence is telling.
Where This Is Heading
The direction of travel is clear, even if the destination is not.
Courts are being asked to confront a new kind of harm, one that does not arise from a defective object or a discrete act, but from a system designed to influence behaviour at scale. The legal system is not built for that kind of problem. It moves incrementally, and it resists sweeping doctrinal change.
If liability is ultimately recognized, I don’t think it will come in a single dramatic judgment. It will emerge slowly, through carefully circumscribed findings tied to specific facts, likely involving minors, demonstrable psychological harm, and internal evidence of deliberate design choices.
Until then, the rhetoric will continue to outpace the law.
The Bottom Line
The idea that social media companies are already being held liable for addictive design is, at best, premature. The cases exist. The arguments are being made. But the legal foundation is still under construction.
For Canadian practitioners, the takeaway is straightforward. Watch the U.S. litigation closely, but do not assume it translates. The doctrinal hurdles here are real, and the courts will not abandon them lightly.
This is not a settled area of law. It is a developing one and it will take time before the courts decide just how far they are prepared to go.
[1] Nate Raymond, Meta must face youth addiction lawsuit by Massachusetts, court rules, Reuters (10 April 2026), online: https://www.reuters.com/world/meta-must-face-youth-addiction-lawsuit-by-massachusetts-court-rules-2026-04-10/
[2] https://fingfx.thomsonreuters.com/gfx/legaldocs/mopaoyzqlpa/04102026meta.pdf
[3] https://cdn.ca9.uscourts.gov/datastore/opinions/2021/05/04/20-55295.pdf
[4] https://www.congress.gov/crs-product/R46751
[5] https://www.supremecourt.gov/opinions/22pdf/21-1333_6j7a.pdf
[6] https://codes.findlaw.com/us/title-18-crimes-and-criminal-procedure/18-usc-sect-2333/
[7] https://www.supremecourt.gov/DocketPDF/21/21-1496/254512/20230210114952851_2023-02-10%20Taamneh%20Merits%20Reply%20FINAL.pdf
[8] https://cand.uscourts.gov/cases-e-filing/cases/422-md-03047-ygr/re-social-media-adolescent-addictionpersonal-injury-products
[10] https://www.canlii.org/en/ca/scc/doc/2011/2011scc47/2011scc47.html
[11] Cooper v. Hobart, 2001 SCC 79, [2001] 3 SCR 537; Rankin (Rankin’s Garage & Sales) v. J.J., 2018 SCC 19, [2018] 1 SCR 587
[12] R. v. Imperial Tobacco Canada Ltd., 2011 SCC 42, [2011] 3 SCR 45, reinforcing that courts must be cautious where claims would have broad policy or regulatory implications.