In early 2026, Aza Raskin, the designer credited with creating infinite scroll, testified in a New Mexico courtroom that the struggle to put your phone down is not only a matter of willpower. On the other side of the screen, he told the court, are thousands of engineers who have run hundreds of millions of tests using your own psychology to keep you there.
Two juries reached related conclusions within a day of each other. In Los Angeles, a federal jury found Meta and Google negligent for designing platforms that are addictive and harmful to children's mental health, awarding six million dollars. In Santa Fe, a separate jury found that Meta had violated New Mexico's Unfair Practices Act by concealing what it knew about harm to young users and the risks of child exploitation on its platforms, a finding that carried a 375 million dollar penalty. The two cases rested on different legal arguments. Together they signalled the same shift: courts are now willing to treat platform design itself as the thing on trial.
Raskin built infinite scroll in 2006 to solve a pagination problem, optimising purely for ease of use. He has said for years that this was the lesson he took from it, that optimising something for ease of use does not mean best for the user or humanity.
The mechanism is straightforward. Traditional page design forces a natural pause. You reach the bottom, decide to click next, and make a conscious choice to continue. Infinite scroll removes that pause. Content loads before your brain registers that a stopping point existed. Combined with variable reward scheduling, where the next post might be fascinating or forgettable, the result is a feedback loop that research links to compulsive use, anxiety, depression, and disrupted sleep.
What makes this relevant to mental health founders is not the litigation. It is the inheritance problem. Habit tracking built around streaks. Mood logging modelled on social app check-ins with completion percentages. Community features built as feeds. None of these defaults are inevitable. They are simply familiar.
Every one of them borrows from the same behavioural psychology playbook Raskin now regrets: variable rewards, the removal of natural endpoints, urgency manufactured through push notifications. The features feel supportive in a mental health context because the content is therapeutic. The underlying mechanism is the same one a jury has just found a company liable for deploying.
One finding from the research is worth foregrounding. Compulsive use, not total screen time, correlates most strongly with declining mental health outcomes. That distinction matters for how mental health tools are evaluated and built. It means the difference between a tool that genuinely helps and one that quietly harms often comes down to a handful of early architectural decisions, rather than a fundamental tradeoff between engagement and wellbeing.
Internal Meta research, surfaced through litigation and whistleblower disclosures, showed the company was aware of harm to younger users. The industry response, across several large platforms, has tended towards optional screen time reminders bolted onto systems still built for engagement. Mental health founders are in a different position. The architecture is still being decided. The defaults have not yet been locked in by a billion adapted users.
The alternative is not a timer retrofitted onto a system built for engagement. It is designing for natural endpoints from the first wireframe, moments where the user has done what they came to do and the product confirms it and lets them leave. It is measuring success by whether people need the tool less over time rather than more. It is making the mechanism visible. Not because the European Union's Digital Services Act now requires risk assessments for platforms whose architecture could cause psychological harm, though it does, but because in a market full of opaque systems, transparency is a design principle before it is a compliance requirement.
The regulatory direction is worth naming plainly. A jury has now found that a company can be held negligent for engineering compulsive use. The Digital Services Act mandates risk assessments for the largest platforms, and in February 2026 the European Commission preliminarily found TikTok in breach of it over addictive design features including infinite scroll. The proposed Digital Fairness Act would extend similar obligations further. Founders building now are not operating before regulation exists. They are operating before it reaches them. That window is the most valuable design opportunity available, and it closes as the product grows.
For founders still in ideation, the cost of building with autonomy, natural endpoints, and transparency as starting requirements is negligible. The cost of retrofitting them onto a system that was built for engagement is not.
🪺