An $18 billion settlement for addictive social media design: the work for health policy starts now Authors: Ilona Kickbusch, Louise Holly, Barbara Prainsack Published: 2026 On 26 August 2026, Meta agreed to pay up to $18 billion to settle litigation brought by a bipartisan coalition of attorneys general from 29 US states and territories. The settlement is intended to resolve claims that the company designed Facebook and Instagram to be addictive, and thus knowingly accepted harm to the mental health of children and adolescents (Ortutay et al. 2026). The company will pay out this sum over 10 years, funding youth mental health programmes, crisis intervention services, and online safety initiatives in the settling states. More importantly, Meta has announced a range of changes to its products to make social media less harmful to users under 18, including a default 2-h daily time limit, blocked access at night, muted notifications during school hours, hidden likes and reactions, restrictions on ‘extreme’ beauty filters and stronger age-assurance measures. Some of these obligations tighten further if other companies, such as YouTube and TikTok, adopt comparable commitments—a ratchet mechanism explicitly designed to pull the rest of the industry along. For the field of health promotion, this settlement is a case study in what can be achieved when a public health harm is recognized and addressed, ex post, through litigation. Many in public health have referred to the tobacco parallel and draw the line from the tobacco Master Settlement Agreement of 1998 to this one: an industry that knew that its product was harmful, concealed evidence (and actively hindered its production), and was eventually forced to pay by the courts (‘The Master Settlement Agreement and Attorneys General’ n.d.). It is precisely this sequence of knowledge, concealment, and eventual liability rather than prevention, that critics now see repeating in the platforms’ teen-safety settlements. In both cases, legal action played such an important role because the political support for a proactive public health approach was too weak, a point we have frequently drawn attention to (Kickbusch and Holly 2023). But not only the tobacco example, but harms from asbestos or opioids use show why litigation is a poor substitute for regulation: it compensates harm after the fact rather than preventing it. It depends on plaintiffs being able to prove individual causation in a domain where harm is often diffuse and cumulative, and its terms are negotiated case by case rather than set as a binding, generally applicable standard. A public health approach would start from the precautionary principle instead: where a product is used by hundreds of millions of children and adolescents and there is a plausible, evidence-based case for harm, the absence of definitive proof of the precise causal mechanism is not a reason to wait. Concretely, this would mean requiring products designed for or knowingly used by minors to undergo mandatory premarket safety assessment, comparable to how medicines or consumer products are assessed before they reach the public, rather than after harm has accumulated. Instead of courts responding to individual lawsuits, regulators should set—and monitor and revise, if necessary—binding design standards for young users. Risks should be identified and addressed before they produce a decade’s worth of documented harm that is then retrospectively addressed in litigation. Proactive regulation is needed to address the root causes of social media harms, going beyond the remedies announced in the settlement agreement. A 2-h default limit or a muted notification schedule is a mitigation measure, not a structural fix. The actions agreed by Meta focus on restricting children’s exposure to harmful features but do not eliminate them. The agreement does not touch the underlying business model of attention extraction and data monetization that makes such design choices commercially rational in the first place. Despite the settlement, Meta continues to deny any wrongdoing. A public health approach goes beyond tort law and asks different questions: what are the commercial incentives, data flows and design defaults that make platforms population-level exposures, and what governance architecture would prevent the harm? Litigation can force disclosure and impose real costs, but it cannot set an enforceable, prospective standard of care for algorithmic design the way regulation can. If the digital determinants of health are, as this journal’s contributors have argued, global by construction—platforms operate across borders on a single technical architecture—then a settlement bounded by US state jurisdiction is at best a partial and geographically arbitrary intervention. The tobacco analogy again offers a sharper contrast: the WHO Framework Convention on Tobacco Control eventually gave the world a shared, binding, prospective standard (WHO Framework Convention on Tobacco Control n.d.). Nothing comparable yet exists, or is seriously being negotiated, for the design of social media platforms and other technologies used by children everywhere. But there are important messages that health promotion should take from this. First, that litigation of this kind can generate real evidence. The internal documents disclosed in discovery are already a public health resource in their own right, much as tobacco industry documents are. Second, that the settlement’s specific defaults (time limits, night-time blocking, hidden metrics, and stronger age assurance) are a reasonable minimum floor that health authorities and school systems can now point to and ask platforms operating in their own jurisdictions to match, regardless of where the litigation happened. Third, and most importantly, that the responsibility for converting this settlement into durable prevention belongs to public health law, to international regulatory cooperation, and to a public health approach that treats platform design as an upstream, modifiable exposure. This needs to be on the agenda of all public health and health promotion associations, conferences, and organizations. And they must speak out! The recent op-ed by the Director General of WHO together with President Macron of France highlighted the challenge (Macron and Ghebreyesus 2026). It is time for the member states of the WHO—especially those already engaged in legislation—to take responsible action through its governing bodies.