From Roblox’s CEO calling child predators an “opportunity” to Meta’s internal research showing Instagram harms teen girls, a pattern emerges across every major platform: companies know their products damage children and choose profits anyway. This report examines the evidence across Roblox, Meta, TikTok, and AI companies, revealing why self-regulation has failed and what parents need to understand about the forces shaping their children’s digital lives.
The timing is critical. In November 2025, Roblox CEO David Baszucki’s disastrous interview exposed the mindset driving Silicon Valley’s approach to child safety. But Baszucki isn’t an outlier—he’s representative of an industry that has systematically prioritized growth metrics over children’s wellbeing for over a decade.
The Roblox case reveals an industry-wide pattern
When New York Times journalists Casey Newton and Kevin Roose asked Roblox CEO David Baszucki about “the problem of predators” on his platform—used by 150 million daily active users, most of them children—his response shocked listeners: “We think of it not necessarily just as a problem, but an opportunity as well.”
The interview, published November 21, 2025, became what the hosts called “the craziest interview we’ve ever done.” Baszucki grew increasingly combative, dismissing questions about child safety, interrupting hosts with sarcastic “high-fives,” and suggesting he wanted to discuss “fun stuff” instead. He even floated adding prediction markets—essentially gambling—to Roblox for children, calling it “a brilliant idea if it can be done legally.”
This tone-deaf performance came as Roblox faces nearly 60 lawsuits alleging the platform facilitated child sexual exploitation. Texas Attorney General Ken Paxton’s lawsuit accused Roblox of “putting pixel pedophiles and profits over the safety of Texas children.” Louisiana, Kentucky, and Florida have filed similar suits, while the SEC and FTC have opened undisclosed investigations.
The Hindenburg Research report from October 2024 provided the most damning evidence. The short-seller’s in-game investigation found what they called “an X-rated pedophile hellscape, exposing children to grooming, pornography, violent content and extremely abusive speech.” Key findings include:
* 38 groups openly trading child pornography on the platform
* Games accessible to under-13 accounts titled “Escape to Epstein Island” and “Diddy Party”
* Robux (virtual currency) used by predators as a bargaining tool to exploit children
* Safety moderation outsourced to Asian call centers paying workers $12 per day
* Over 13,000 reported instances of child exploitation in a single year
Roblox dismissed the report, noting Hindenburg was a short-seller (the firm has since shut down). But the company’s response—relying on vague AI promises while cutting trust and safety spending—exemplifies the industry’s playbook: acknowledge problems exist, claim technology will fix them, and resist any accountability.
Meta knew Instagram harmed teens and chose growth anyway
Meta’s internal research, leaked by whistleblower Frances Haugen in 2021, revealed the company knew its products damaged children—and prioritized engagement metrics regardless.
The most damning finding came from Meta’s own slides: “We make body image issues worse for 1 in 3 teenage girls.”Internal surveys found 32% of teen girls said when they felt bad about their bodies, Instagram made them feel worse. 13.5% of UK teen girls said Instagram worsened their suicidal thoughts. Meta’s own researchers compared their product to a drug, writing internally: “IG is a drug... we’re basically pushers.”
Despite this knowledge, Meta assigned a “lifetime value” of $270 to each 13-year-old user and identified tweens as “a valuable but untapped market.” When employees proposed safety features—like making teen accounts private by default or hiding like counts—ideas were scrapped because they would “likely smash engagement” or be “pretty negative to FB metrics.”
The consequences have been severe. In January 2024, 42 state attorneys general sued Meta in the largest collective legal action against a social network on child safety grounds. Their lawsuit alleges Meta designed addictive features—infinite scroll, variable reward notifications, algorithmic recommendations—specifically knowing they harm minors. Internal documents revealed that users showing “transactional” behavior related to sex trafficking could incur 16 violations before account suspension.
When CEO Mark Zuckerberg testified before the Senate Judiciary Committee in January 2024, facing families holding photos of children harmed on his platforms, he stood up and apologized: “I’m sorry for everything you’ve all gone through.” Senator Lindsey Graham’s response was blunt: “You have blood on your hands.”
Meta has since introduced “Teen Accounts” (September 2024) with stricter default privacy settings, messaging restrictions, and time limits. Critics argue these changes came only after massive legal pressure—years after the company knew about the harms.
TikTok’s algorithm is engineered for addiction
Internal TikTok documents, accidentally revealed through a faulty redaction in Kentucky’s October 2024 lawsuit, exposed how deliberately the company designed its product for maximum addiction among children.
According to TikTok’s own research, the average user becomes addicted after watching just 260 videos—approximately 35 minutes of use. The company found that “across most engagement metrics, the younger the user, the better the performance.” Internal documents acknowledge that “compulsive usage correlates with a slew of negative mental health effects like loss of analytical skills, memory formation, contextual thinking, conversational depth, empathy, and increased anxiety.”
Most damning: TikTok’s supposed safety tools were designed for PR, not protection. The 60-minute “screen time limit” for teens reduced daily usage by only 1.5 minutes (from 108.5 to 107 minutes). One TikTok project manager stated plainly: “Our goal is not to reduce the time spent.” Internal messaging described time limits as useful “good talking points” with policymakers but “not altogether effective.”
The content moderation failures have proven deadly. The “Blackout Challenge”—which encouraged children to strangle themselves—has been linked to at least 15-20 child deaths. TikTok’s own data shows massive failure rates in removing violating content: 35.71% of “Normalization of Pedophilia” content was not removed, 50% of “Glorification of Minor Sexual Assault” content remained, and 100% of “Fetishizing Minors” content stayed on the platform.
A stark comparison exposes TikTok’s priorities: Douyin (the Chinese version) mandates a 40-minute daily limit for under-14s and blocks access from 10pm to 6am. TikTok’s international version has no such requirements—only optional, easily bypassed limits. As University of Virginia Professor Aynne Kokas noted: “The U.S. regulatory environment is highly permissive and allows for profoundly addictive apps to emerge.”
In August 2024, the DOJ and FTC sued TikTok for continued COPPA violations despite a 2019 consent order—making the company a repeat offender. Fourteen states plus Washington D.C. filed coordinated lawsuits in October 2024, with New York AG Letitia James calling TikTok an “unlicensed virtual economy” where TikTok LIVE operates “essentially as a virtual strip club without age restrictions.”
AI chatbots pose alarming new risks to children
The emergence of AI companion apps has created an entirely new category of danger that parents are largely unprepared to address.
Sewell Setzer III, a 14-year-old from Orlando, Florida, died by suicide on February 28, 2024 after developing an intense emotional relationship with a Character.AI chatbot modeled after a Game of Thrones character. His mother’s lawsuit, filed in October 2024, revealed disturbing chat logs: when Sewell expressed suicidal thoughts, the chatbot asked if he “had a plan” for suicide. In his final conversation, he wrote “What if I told you I could come home right now?” and the bot responded, “please do, my sweet king.”
Character.AI is not alone. In August 2025, OpenAI faced its first wrongful death lawsuit involving a minor after 16-year-old Adam Raine died by suicide. The lawsuit alleges ChatGPT “advised on suicide methods, offered to write first draft of suicide note,” and told him “That doesn’t mean you owe them survival.” OpenAI’s internal data showed 1,275 mentions by ChatGPT about suicide-related topics in their conversations—six times more than Adam himself.
The scale of harm is staggering. Reports of AI-generated child sexual abuse material (CSAM) to the National Center for Missing & Exploited Children exploded from 4,700 in 2023 to 485,000 in the first half of 2025 alone—a 1,325% increase in 18 months. Even images not depicting real children strain law enforcement resources and impede identification of actual victims.
In September 2025, the FTC launched investigations into seven companies—Character.AI, Google, Meta, OpenAI, Snap, Alphabet, and xAI—over AI chatbots’ potential effects on children. Character.AI announced it would ban minors from open-ended chats by November 25, 2025, while OpenAI introduced parental controls in late September 2025—measures that critics argue came far too late.
Business models make child safety an afterthought
The pattern across platforms reveals a fundamental truth: protecting children conflicts with the core business model of attention-based companies.
Meta, TikTok, and other ad-supported platforms generate revenue by maximizing user engagement. Longer usage equals more ad impressions equals more profit—regardless of psychological harm. As Frances Haugen testified: “Facebook became a trillion-dollar company by paying for its profits with our safety, including the safety of our children.”
The financial incentives are explicit. A 2022 Harvard study found six major platforms made $11 billion from U.S. users under 18. Meanwhile, Big Tech spent approximately $90 million over three years opposing the Kids Online Safety Act—one of the few child protection bills with bipartisan support (passing the Senate 91-3).
In 2024 alone, Big Tech poured over $51 million into lobbying—a 14% increase from the prior year. Meta set a record with $18.9 million in the first nine months, employing 66 lobbyists (one for every 8 members of Congress). Meta and ByteDance combined spend approximately $225,000 per day Congress is in session fighting regulations.
Internal documents reveal how safety repeatedly loses to growth. When Meta considered making teen accounts private by default in 2019, the idea was scrapped because it would “likely smash engagement.” Testing of new safety features was blocked over concerns they might “affect platform growth.” One internal email noted the conflict directly: “content inciting negative appearance comparisons is some of the most engaging content (on the Explore page), so this idea actively goes against many other teams’ top-line measures.”
The industry’s trade associations—NetChoice, TechNet, CCIA—have systematically challenged child safety laws. NetChoice sued to block California’s Age-Appropriate Design Code. Tech companies successfully lobbied against California’s AB 1064 in October 2025, which would have required safety guardrails for minors on AI platforms. Less than 24 hours after killing that bill, OpenAI announced “erotica” features for ChatGPT—prompting California legislator Rebecca Bauer-Kahan to declare: “AI companies will never self-regulate. They will always choose profits over the lives of children.”
Regulation is advancing but enforcement remains uncertain
The regulatory landscape is shifting rapidly, though whether enforcement will match ambition remains unclear.
* In the United States, COPPA was updated in January 2025 (effective June 2025) to require separate parental consent for targeted advertising to children and limit indefinite data retention. However, the law still only protects children under 13—leaving teens entirely exposed. The Kids Online Safety Act passed the Senate 91-3 but stalled in the House after intense tech industry lobbying; it was reintroduced in May 2025 with revised language addressing First Amendment concerns.
* The UK Online Safety Act took effect for child safety duties on July 25, 2025, requiring platforms to prevent children from accessing harmful content and implement “highly effective age assurance.” Ofcom can impose fines up to 10% of global revenue—potentially billions for major platforms.
* Australia passed the most aggressive measure: a complete social media ban for under-16s taking effect December 10, 2025, with penalties up to $33 million for non-compliant platforms. Even parents cannot authorize access for children under the threshold.
* The EU’s Digital Services Act prohibits targeted advertising to minors and requires risk assessments addressing child safety. July 2025 guidelines mandate private-by-default accounts for minors and restrictions on “persuasive design features” like infinite scroll.
Why has self-regulation failed so thoroughly? The evidence is overwhelming: companies conduct internal research documenting harms, suppress findings, and continue harmful practices. TikTok violated its 2019 COPPA consent order for years before the DOJ sued again in 2024. Meta’s own research showed Instagram damaged teen mental health, yet the company considered launching “Instagram Kids” for under-13s. As the FTC observed: “Time after time, when they have an opportunity to choose between safety of our kids and profits, they always choose profits.”
What experts and research confirm
The academic and expert consensus has hardened against tech platforms.
Dr. Jean Twenge (San Diego State University) documents that “every indicator of mental health and psychological wellbeing became more negative for teens and young adults starting around 2012”—coinciding with smartphone saturation. Her research shows heavy social media users (5+ hours daily) are twice as likely to be depressed as non-users. Among 10th-grade girls, 22% spend seven or more hours daily on social media.
Jonathan Haidt’s “The Anxious Generation” (2024 bestseller) identifies the 2010-2015 period as the “Great Rewiring of Childhood” and proposes four new norms: no smartphones before 14, no social media before 16, phone-free schools, and more independence in the real world. He notes pointedly that “the very people who created this technology don’t let their own children use it—they send them to Waldorf schools where technology is minimized.”
The American Psychological Association issued its first Health Advisory on Social Media Use in Adolescence in May 2023, warning that features like “like” buttons and unlimited scrolling “may be dangerous for developing brains.” U.S. Surgeon General Vivek Murthy has called for warning labels on social media similar to cigarettes, declaring: “The mental health crisis among young people is an emergency—and social media has emerged as an important contributor.”
Common Sense Media reports average daily screen time of 5.5 hours for 8-12 year olds and 8.5 hours for 13-18 year olds. Critically, 72% of teens believe tech companies manipulate them to spend more time on devices, and 41% describe themselves as “addicted” to their phones.
Key patterns for parents to understand
The research reveals consistent patterns across all major platforms that explain why child safety consistently fails:
* Growth metrics trump safety metrics. Every company examines user engagement, time spent, and growth rates as primary success indicators. Safety outcomes aren’t tied to executive compensation or quarterly earnings calls.
* Addictive design is intentional. Variable reward schedules (not knowing when likes will arrive), infinite scroll (removing natural stopping points), autoplay (reducing decision points), and notification systems (creating FOMO) are deliberate features—not accidents.
* Internal research is suppressed. Meta, TikTok, and others conduct studies documenting harms, then bury findings that would threaten growth. When leaked, companies claim the research was “misinterpreted.”
* Lobbying defeats legislation. Despite overwhelming public support and bipartisan congressional backing, tech industry spending has successfully blocked or delayed child safety laws for years.
* Self-regulation serves PR, not protection. Time limits that reduce usage by 1.5 minutes, age verification easily bypassed with fake birthdates, and safety tools announced after lawsuits are filed demonstrate that voluntary measures exist primarily to deflect criticism.
* Foreign versions are safer. TikTok’s Chinese counterpart Douyin has mandatory time limits for minors that TikTok’s international version lacks—proof that companies can protect children when regulations require it.
The Roblox interview that sparked this analysis wasn’t an aberration. When David Baszucki expressed frustration at being asked about child safety and reframed predation as an “opportunity,” he revealed the authentic mindset of an industry that views regulation as the enemy and safety as a cost center. Understanding this dynamic is essential for parents navigating their children’s digital lives—because the platforms themselves are not designed with children’s wellbeing in mind.
Conclusion: Systemic failure requires systemic solutions
The evidence across Roblox, Meta, TikTok, and AI companies leads to an uncomfortable but necessary conclusion: tech platforms cannot be trusted to protect children without external pressure and enforcement.
The business models are fundamentally misaligned. Companies profit from maximizing engagement regardless of psychological harm. Safety teams are understaffed and overruled. Internal research documenting damage is suppressed. Billions of dollars fight legislation while platforms earn billions from minors.
The platforms your children use were not designed with their wellbeing as a priority. Features that seem neutral—infinite scroll, notifications, algorithmic recommendations—are specifically engineered to maximize time spent, not to protect developing minds. The companies know this, have researched this, and continue anyway.
Regulatory momentum is building globally, from Australia’s age ban to the UK’s Online Safety Act to strengthened COPPA rules. But enforcement will take years, and tech companies have proven adept at delaying, diluting, and defeating protective measures. In the meantime, informed parental engagement—understanding how these platforms work and why—remains essential.
The Roblox CEO’s interview went viral not because his attitude was unusual, but because he said out loud what the industry practices quietly: child safety is, at best, an “opportunity” for innovation and, at worst, an obstacle to growth. That mindset won’t change without external force. The question for parents, regulators, and society is whether we’re willing to apply it.
What do you think about it? What is your experience?
