Meta Platforms has agreed to a settlement worth up to $18 billion over the next decade to resolve child safety allegations brought by a coalition of American states. Yet the agreement is drawing sharp criticism from Arturo Bejar, the former Meta safety engineer whose testimony proved instrumental in building the states' case, who argues that the deal fails to confront the core issues he raised in court regarding how the company's platforms harm young users.
Bejar's courtroom statements detailed a troubling pattern: Meta knowingly downplayed how frequently teenagers experienced harm across Facebook and Instagram, deliberately promoted content designed to generate negative feelings about their appearance and bodies, and inadequately investigated reports of child sexual exploitation. These allegations struck at the heart of years-long public controversy over whether social media platforms prioritise user engagement over youth wellbeing.
California's Attorney General Rob Bonta hailed the settlement as a watershed moment that would "make social media less dangerous for our kids," framing it as a vindication of the states' aggressive enforcement stance. Yet Bejar took an opposing view, characterising the agreement as likely to entrench what he called "safety theater"—the appearance of reform without meaningful change. He suggested that while the measures might reduce time spent on Instagram, they would not genuinely improve the safety experience for young users engaging with the platform.
The settlement requires Meta to implement several safeguards, including expanded parental monitoring tools, daily time limits for teenage users, and measures to detect minors falsifying their age. Meta also agreed to hide like and reaction counts on posts to counteract what researchers term "negative social comparison," where teenagers measure their social worth against peers' post performance. This particular remedy illustrates a broader tension at the heart of the case: internal Meta research dating to 2019 showed that hiding likes had only minimal impact on user wellbeing, and the company calculated that implementing the change would reduce daily active users by approximately 0.09%—a negligible business consequence.
When Meta tested the likes-hiding feature in an initiative known as "Project Daisy," employees presented findings to CEO Mark Zuckerberg showing that overall wellbeing metrics barely moved. Despite this evidence that the measure would solve little, it took external legal pressure for Meta to even offer it as an optional feature. The settlement now mandates something the company itself determined would have only marginal benefits, raising uncomfortable questions about whether litigation-driven remedies can effectively address structural design problems.
Internally leaked research over the years has established that Meta's algorithms disproportionately recommend fitness and beauty content to teenagers already struggling with self-esteem, and that the platform's architecture deliberately extends user engagement beyond what people would voluntarily choose. These design choices embed harms into the product itself. Bejar's own work revealed that Meta's official statistics on user harm rates substantially understated actual prevalence, suggesting either systematic underestimation or deliberate underreporting.
The settlement addresses these algorithmic and design concerns only peripherally, instead directing attention toward parental controls and time restrictions. This mismatch between the substance of Bejar's allegations and the scope of the remedies suggests that the settlement may represent a compromise that satisfies political demand for action without requiring fundamental changes to Meta's business model. The emphasis on parental oversight, while superficially reasonable, also transfers responsibility for policing teen behaviour away from the platform itself.
Meta's capacity to enforce age-based restrictions remains questionable. The company relies on user self-reporting for age verification, a system vulnerable to falsification. In Australia, where the government implemented an outright ban on social media use by teenagers under sixteen, authorities discovered that eight in ten young teens continue accessing platforms despite the prohibition—underscoring how difficult meaningful age-based enforcement proves in practice.
Mental health experts expressed cautiously divided opinions about the settlement's significance. Dr. Jane Conron, a clinical psychologist at Northwestern University's Feinberg School of Medicine, noted that the opt-in non-algorithmic feed will likely attract minimal uptake, as most users default to whatever settings the platform establishes. However, she saw potential value in the daily usage caps, which some of her adolescent patients find emotionally distressing to navigate independently. She described instances where teenagers experience such acute distress at the prospect of disconnection that parents encounter resistance bordering on crisis when attempting to enforce limits.
Conron suggested that the settlement's most meaningful impact might be indirect. By formally acknowledging the gravity of social media's effects on youth mental health, the agreement potentially shifts cultural and family conversations around device use. Rather than dismissing parental concerns as generational anxiety, the Meta settlement—regardless of its specific remedies—validates the underlying worry that these platforms affect young people's psychological wellbeing in consequential ways.
The settlement notably stops short of requiring Meta to admit fault or concede that its products were deliberately constructed to cause harm. This legal protective measure allows the company to maintain its public positioning while acceding to remedies. For policymakers and regulators observing from markets like Malaysia and Southeast Asia, where social media penetration among youth remains extremely high and regulatory frameworks remain in development, the Meta case illustrates both the power and the limitations of post-hoc litigation. The settlement demonstrates that sufficient public pressure and legal coordination can extract concessions from technology giants, yet also reveals how difficult it remains to translate legal victories into fundamentally safer products when the business incentives pulling toward engagement-maximisation remain structurally intact.
