Meta Platforms' $18 billion settlement with a coalition of American states has drawn sharp criticism from Arturo Bejar, the former safety engineer whose damning testimony was instrumental in building the child safety case against the social media giant. Speaking publicly after the deal was announced, Bejar contended that while the settlement may appear comprehensive on the surface, it largely sidesteps the substantive problems he revealed during court proceedings—namely that Meta deliberately misreported how frequently teenagers faced harm on its platforms, algorithmically steered young users toward content designed to make them insecure, and failed adequately to investigate reports of child predation.
The settlement represents a significant enforcement action, with Meta committing to pay billions over the coming decade and accepting various restrictions on how teenagers access Facebook and Instagram. California Attorney General Rob Bonta heralded the agreement as a watershed moment, declaring it would genuinely reduce dangers facing young people online. Yet Bejar's critique cuts to the heart of a deeper tension: the measures Meta has accepted, while costly in absolute terms, may represent little more than symbolic gestures that do nothing to alter the fundamental architecture through which the platforms operate and profit.
Central to Bejar's scepticism is the observation that several remedies embedded in the settlement were previously considered by Meta's own internal teams and consciously abandoned because they posed minimal threat to either user engagement or corporate interests. The most telling example involves Meta's commitment to hide likes and reaction counts on posts—a feature designed to reduce what researchers term "negative social comparison" among adolescents. This concept was explored as far back as 2019, when Meta initiated an internal investigation into whether hiding such metrics might alleviate the documented psychological harms teenagers experience through constant peer evaluation on social media.
Meta's internal testing, known as Project Daisy, produced results that proved disappointing from a mental health perspective but reassuring from a business standpoint. Internal presentation slides circulated to Mark Zuckerberg in 2020 revealed that hiding like counts produced only marginal improvements in overall well-being measures among users. Simultaneously, modelling suggested that adopting the change permanently might suppress daily active users by a mere 0.09 percent—economically negligible for a platform valued in the hundreds of billions. Rather than implementing the measure universally, Meta chose to offer it as an opt-in feature, essentially burying it from view and rendering it practically irrelevant to most teenage users.
The settlement's shortcomings extend beyond individual features to encompass Meta's fundamental business model. Bejar and other researchers have documented how Meta's algorithms are deliberately engineered to recommend high-intensity content related to fitness and beauty ideals specifically to teenagers already struggling with self-esteem. The platforms employ sophisticated design techniques—infinite scrolling, variable reward schedules, colour psychology—explicitly intended to extend user session durations beyond what young people would naturally select. These design choices are not incidental; they are foundational to how Meta maximizes engagement metrics and, consequently, advertising revenue.
The settlement addresses these algorithmic and design concerns only peripherally. Instead, the remedies emphasize parental oversight mechanisms, usage-time restrictions, and enhanced capacity to identify minors falsely claiming to be adults. Each of these measures places responsibility squarely on parents and individual users rather than requiring Meta to fundamentally alter how its platforms function. The assumption underlying this approach—that the problem lies primarily in how teenagers access these services rather than in the services themselves—represents a convenient framing for Meta but a misdiagnosis of the underlying pathology.
Meta's capacity to enforce even the settlement's own terms remains questionable. The company's age-verification mechanisms depend substantially on users self-reporting their birthdates—a system vulnerable to circumvention by any teenager with basic digital literacy. Recent experience in Australia illustrates this vulnerability starkly. When the Australian government prohibited social media access for users under sixteen, regulators subsequently discovered that approximately eight in ten young teenagers remained active on these platforms despite the formal ban. This finding suggests that compliance-based approaches, no matter how sophisticated their nominal requirements, will likely prove porous in practice.
Expert reaction among clinical psychologists and adolescent health specialists has been decidedly mixed. Dr. Jane Conron, a clinical psychologist at Northwestern University's Feinberg School of Medicine, identified one potentially valuable element: the settlement's requirement that Meta implement built-in daily time limits on Instagram and Facebook usage. Some of Conron's young patients experience such acute emotional distress when separated from these applications that they weep when parents attempt to enforce restrictions. A system-level usage cap might provide meaningful relief for such individuals and, perhaps more importantly, shift how families discuss social media consumption.
However, Conron expressed significant reservations about other settlement provisions. The requirement that Meta offer a non-algorithmic feed option to teenagers proves nearly worthless, she suggested, because the opt-in mechanism virtually guarantees that teenagers—precisely the demographic most susceptible to algorithmic manipulation—will not elect to use it. Recognising that most users will select whatever default setting Meta provides, the settlement effectively permits the company to maintain algorithmic curation as the dominant experience.
The settlement's psychological impact may ultimately prove more significant than its mechanical requirements. Conron observed that even without Meta's explicit acceptance of wrongdoing, the very existence of this enforcement action signals that institutions and authorities now take adolescent mental health risks seriously. This recognition, she suggested, might catalyse broader cultural conversations between parents, teenagers, educators, and policymakers about appropriate social media use and the genuine hazards these platforms pose. The settlement, viewed through this lens, becomes less a solution to structural problems and more a opening for genuine reform—though only if stakeholders move beyond the agreement's insufficient minimum standards.
For Malaysian and Southeast Asian contexts, this settlement carries indirect but significant implications. Meta's platforms, particularly Facebook and Instagram, dominate social media usage across the region, with Malaysia's teenage population representing a particularly engaged user base. Whatever protections or safeguards Meta implements globally tend to roll out across all markets simultaneously. Conversely, the company's demonstrated willingness to treat mental health protections as negotiable business expenses rather than non-negotiable rights suggests that regional regulators cannot rely on Meta's voluntary compliance or American legal settlements to adequately protect young users in their own jurisdictions. The lesson appears clear: meaningful protection requires proactive regional intervention.
