When a blockbuster trial opened this month in California, four US states sued Meta over allegations that Facebook and Instagram designed addictive features harming children's mental health. Few expected the case to resolve in days, but last week Meta agreed to pay up to $18bn, one of the largest settlements ever reached by a US tech company. Meta also committed to safeguards intended to make its platforms safer for young users.

While they do not yet go far enough, they could be expanded internationally. Trials are not the best way to resolve public safety issues, but US litigation has again shown its capacity to force changes in corporate behaviour where legislators and regulators have failed. Even $18bn is small change for Meta, well below the $200bn the states had sought.

Meta said payment of $5.3bn of the total is contingent on YouTube and TikTok each agreeing matching payments and imposing similar restrictions on teens. For now, the safety commitments apply only to Meta and only in the US. The attempt to bind YouTube and TikTok is significant, as research suggests teens use their platforms more frequently, and both face litigation too.

The settlement requires Meta to build an age-assurance framework to tighten the age bar on under-13s and distinguish 13 to 17-year-olds from adults, using either third-party or its own technology. Performance standards include annual third-party testing. The strongest parts of the deal are a two-hour limit on use across Facebook and Instagram, a blackout period from midnight to 6am, and a block on push notifications from 10pm to 7am and during school hours. Similarly welcome is hiding "likes" and reaction counts by default on posts created by children's accounts, and banning "beauty" filters that simulate cosmetic enhancements, which can intensify body image anxiety.