The sweeping $18 billion settlement between Meta Platforms and a coalition of US states has drawn sharp criticism from the former safety engineer whose explosive testimony helped propel the child protection case against the social media giant. Arturo Bejar, who previously worked as a safety specialist at Meta, disputed claims that the agreement meaningfully tackles the fundamental problems he described during his court testimony last week. The whistleblower's concerns centre on whether the settlement's proposed remedies will actually translate into safer platforms for young users or merely represent what he characterises as additional layers of corporate window-dressing.
During his testimony, Bejar presented evidence that Meta systematically underreported the extent of harm experienced by teenagers using its services. He detailed how the company's algorithms deliberately steered young users toward content known to damage their self-image and mental wellbeing, and how Meta's systems failed to adequately protect minors from predatory behaviour. These allegations formed the backbone of the states' case against the technology firm, with California Attorney General Rob Bonta announcing the settlement by claiming it would "make social media less dangerous for our kids." Yet Bejar's immediate response suggested the deal fell far short of addressing his core concerns about how Meta operates its platforms.
The settlement requires Meta to pay states up to $18 billion over the next decade and implement various protective measures targeting teenage users on Facebook and Instagram. However, Bejar argued that many of these remedies represent measures the company had previously evaluated and deliberately rejected as ineffective. This revelation raises uncomfortable questions about whether regulators have truly extracted meaningful concessions or simply negotiated the adoption of safeguards that Meta's own internal research had determined would have minimal impact on user wellbeing. The distinction matters significantly for Southeast Asian regulators watching how other democracies handle tech platform oversight.
One prominent example involves Meta's agreement to hide the likes and reaction counts visible on user posts. The company claims this addresses concerns about negative social comparison among young people, yet the origins of this remedy trace back to 2019 when Meta researchers were investigating whether social media contributed to teenage mental health problems. Meta's internal testing, documented in presentations to Mark Zuckerberg, found that hiding like counts produced at most modest effects on user experience. The company's own employees acknowledged in 2020 slide presentations that the change did not measurably shift overall wellbeing metrics among users.
Meta's calculus extended beyond wellbeing considerations to business impact. Testing conducted through a project codenamed "Project Daisy" indicated that adopting the like-hiding measure as a default would result in only a marginal 0.09 percent decline in daily active users. This minimal projected revenue impact made the measure acceptable for inclusion in the settlement, yet the company had previously resisted implementing it as a standard feature, choosing instead to offer it as an optional user setting. The pattern suggests regulators may have negotiated agreements that allow Meta to demonstrate compliance with relatively superficial changes that the company had already determined posed negligible business risks.
Bejar's broader concern transcends any single feature. He characterises the settlement as codifying what he describes as "safety theater," implying that while Instagram usage might decline slightly, the platforms themselves will not fundamentally become safer spaces for teenagers. This distinction between reduced usage and reduced harm cuts to the heart of regulatory effectiveness. Parents might feel reassured that their children spend less time scrolling, yet if the time spent remains psychologically damaging, the settlement may provide only the illusion of protection rather than substantive improvements to platform design.
The settlement's architecture reveals additional gaps between stated objectives and concrete remedies. While the states' case centred on Meta's algorithmic recommendations, content moderation failures, and underreporting of harm prevalence, the settlement emphasises enhanced parental controls, time limits, and improved detection of age-misrepresenting users. This pivot toward parental oversight rather than platform redesign essentially shifts responsibility for protecting teenagers from Meta to individual families. Yet Meta's fundamental ability to verify user age remains questionable, relying on self-reported information that users routinely falsify.
Evidence from Australia illustrates the practical limitations of age-based restrictions in the social media context. After implementing a comprehensive ban on social media for users under 16, Australian regulators discovered that approximately 80 percent of young teenagers remain active on platforms. This discrepancy between regulatory intent and actual compliance suggests that approaches emphasising user verification and age gates face inherent implementation challenges. For Malaysian policymakers contemplating similar protections, the Australian experience provides cautionary lessons about the gap between legislated restrictions and their real-world enforcement.
Expert reaction to the settlement has remained divided, reflecting genuine uncertainty about its practical utility. Dr. Jane Conron, a clinical psychologist at Northwestern University's Feinberg School of Medicine, expressed scepticism about the settlement's most prominent feature: an option for teenagers to access algorithmically unfiltered feeds. Conron predicted this opt-in approach would attract minimal usage among teens, who would likely continue relying on algorithm-driven recommendations that optimise for engagement rather than wellbeing. However, she identified potential value in the settlement's provisions requiring Meta to implement daily usage caps on Instagram and Facebook for teenage users.
Conron's clinical experience suggests that usage limits could provide meaningful intervention points for families struggling with adolescent screen addiction. Several of her young patients have demonstrated such intense emotional dependence on social media that they become visibly distressed when parents attempt to restrict access. Built-in technical limits bypass the need for parent-child negotiation around usage time, potentially reducing household conflict while protecting vulnerable teenagers from their own compulsive behaviour. Additionally, Conron suggested that the settlement's existence itself might shift cultural conversations about social media's role in teenage life, even without specific acknowledgment of platform culpability.
Notably, the settlement required no admission of fault from Meta and does not establish that the company deliberately built harmful products. This structural limitation preserves Meta's legal position while allowing regulators to claim victory through the implementation of specific remedies. For international observers, including policymakers in Malaysia and Southeast Asia, the settlement illustrates how regulatory negotiations can produce significant financial penalties and operational changes without establishing clear causal links between platform design and documented harms. The absence of acknowledged wrongdoing complicates future enforcement efforts and may embolden platforms to argue that similar harms in other jurisdictions result from extraneous factors rather than product design choices.
The gap between Bejar's testimony and the settlement's scope ultimately raises questions about regulatory capture and the adequacy of current legal frameworks for addressing algorithmic harms. When a company's own internal research identifies specific design choices that damage teenage users, and when a whistleblower credibly testifies to these documented harms, a settlement that fails to mandate fundamental algorithmic redesign suggests that existing regulatory approaches may be insufficient. For Southeast Asian regulators developing their own social media oversight strategies, the Meta settlement demonstrates both the potential and limitations of using financial penalties and feature-level modifications as tools for protecting young users from platform harms rooted in business model incentives.
