California AG Expands Teen Safety Probe to TikTok, YouTube

California's attorney general has launched formal investigations into TikTok and YouTube over alleged harms to minors, expanding a sweeping state-level push into platform accountability that previously centred on Meta's record $1.4 billion multistate settlement. The probe signals a significant escalation in regulatory pressure on the social media industry at a moment when international bodies, including the United Nations, are also calling for binding protections for young users online.

The Scope of the California Investigation

Attorney General Rob Bonta's office confirmed the formal expansion of its youth safety probe to include ByteDance-owned TikTok and Alphabet's YouTube, officials said. The investigations follow complaints that both platforms employ recommendation algorithms — the automated systems that decide which videos and posts appear in a user's feed — designed in ways that may exploit adolescent psychology to maximise engagement, even when that engagement involves harmful content.

Recommendation algorithms work by analysing a user's behaviour: what they watch, how long they linger, what they share. Over time, these systems build a detailed profile and serve increasingly targeted content. Critics argue that, for teenagers whose brains are still developing impulse control and emotional regulation, this creates a feedback loop that amplifies anxiety, eating disorders, and self-harm ideation. The California AG's office is specifically examining whether both companies violated the state's consumer protection statutes and its recently enacted AADC-aligned child privacy provisions, according to people familiar with the matter.

What the Meta Settlement Established

The framework for these new investigations was shaped significantly by the Meta case, in which a coalition of attorneys general from more than 40 states secured the $1.4 billion agreement earlier this year. That settlement required Meta — the parent company of Instagram and Facebook — to implement independent audits of its youth-facing features and restrict targeted advertising to users under 18. Legal analysts note the Meta case created a replicable enforcement template that state AGs can now apply to other platforms (Source: Reuters).

The California office has drawn on internal documents disclosed during the Meta litigation, including research reportedly showing that Instagram's own teams identified significant mental health impacts on teenage girls. Whether similar internal research exists at TikTok or YouTube, and whether it was withheld from regulators, is a central question in the current probe, officials indicated.

TikTok Under Dual Pressure

TikTok's situation is complicated by the ongoing national security debate in Washington, where legislation compelling ByteDance to divest its U.S. operations has created legal uncertainty. However, consumer protection investigations by state attorneys general operate independently of federal national security reviews, meaning TikTok faces simultaneous legal exposure on multiple fronts.

Content Moderation Practices in the Spotlight

Regulators and researchers have raised persistent concerns about TikTok's For You Page — its primary algorithmic feed — and its capacity to serve vulnerable teenagers content related to weight loss, self-harm, and extremist communities within minutes of account creation. A series of undercover tests conducted by nonprofit researchers and reported in Wired demonstrated that newly created accounts identifying as teenagers were served eating disorder content in under thirty minutes of browsing.

TikTok has introduced features including screen time limits for users under 18, restricted mode settings, and the ability for parents to pair their accounts with their children's through Family Pairing. Company officials maintain these tools represent meaningful safeguards. Regulators and child safety advocates argue the tools are opt-in rather than default, making them largely ineffective for the populations most at risk (Source: Associated Press).

British regulators have pursued a parallel track. For more on how TikTok's moderation practices are being examined in the UK, see our earlier coverage of how UK regulators are probing TikTok's content moderation practices, which offers useful context on the transatlantic dimension of this story.

YouTube's Specific Liability Questions

YouTube's investigation centres on somewhat different concerns. As the dominant video platform globally, YouTube's recommendation engine has been the subject of academic study for years. MIT Technology Review has published extensive research documenting the so-called "rabbit hole effect," in which the recommendation algorithm progressively surfaces more extreme or emotionally intense content to retain viewer attention. For adolescent users, that progression has been linked in some studies to radicalisation pathways and increased consumption of content promoting disordered eating.

Autoplay and the Architecture of Engagement

One specific feature under scrutiny is YouTube's autoplay function, which automatically queues a subsequent video when one ends. While autoplay can be disabled, it is enabled by default. Child safety researchers argue that for a young teenager watching content late at night, autoplay removes the deliberate decision to continue watching, bypassing the moment where a user might otherwise choose to stop. The attorney general's investigation is examining whether this constitutes an unfair or deceptive design practice under California consumer law, officials said.

Google, YouTube's parent company, has said the platform offers supervised accounts for children and has gradually extended age-appropriate content restrictions. A company spokesperson said YouTube is cooperating with regulators and pointed to its existing suite of parental controls as evidence of good faith compliance efforts (Source: Reuters).

The UN and the Push for International Standards

The California investigation does not exist in isolation. The United Nations Special Rapporteur on the Right to Privacy has recently called on member states to adopt binding international standards governing platform design choices that affect minors, citing evidence that voluntary industry self-regulation has been insufficient. The UN's intervention adds diplomatic weight to what has until recently been a fragmented, jurisdiction-by-jurisdiction regulatory effort.

Industry analysts at IDC have noted that the convergence of state, federal, and international pressure represents a structural shift in the regulatory environment for social media companies. Rather than managing isolated legal risks in specific markets, platforms now face overlapping enforcement regimes that, taken together, may compel substantive changes to core product architecture — not merely surface-level policy adjustments (Source: IDC).

Legislative Momentum in Key Markets

Legislators in the European Union have already moved to enforce the Digital Services Act's provisions on algorithmic transparency, requiring platforms above a certain size threshold to submit to independent algorithmic audits. In the United Kingdom, amendments strengthening children's protections under the Online Safety framework continue to advance. Readers seeking background on the UK legislative context can follow the progress of UK Parliament's Online Safety Bill amendments, which directly address platform duties of care toward minors.

The pace of legislative change has also accelerated AI-specific guardrails. Separately, the AI industry is grappling with its own obligations toward younger users, a discussion covered in our analysis of how OpenAI's teen safety measures are setting new standards for U.S. AI guardrails — relevant context as AI-driven recommendation systems increasingly blur the line between traditional social media and generative AI products.

Industry Response and the Limits of Self-Regulation

Both TikTok and YouTube have historically argued that their platforms provide significant educational, creative, and social value to teenagers, and that regulatory intervention risks curtailing free expression. Neither company has issued a detailed public response specifically addressing the California AG's expanded probe as of the time of publication.

The broader debate about self-regulation versus statutory obligation is not new, but the scale of the Meta settlement has fundamentally altered the calculus for platform legal teams. Gartner analysts have observed that social media companies are now allocating substantially larger portions of their legal and trust-and-safety budgets to proactive regulatory engagement, anticipating that voluntary concessions made early in an investigation can reduce ultimate financial exposure (Source: Gartner).

What Comes Next

Legal experts tracking the California cases say the attorney general's office is likely to issue formal civil investigative demands — the equivalent of subpoenas in a civil proceeding — requiring TikTok and YouTube to produce internal research, communications, and algorithmic documentation relating to teenage users. Both companies will face pressure to demonstrate that their design choices reflect genuine safety considerations rather than engagement optimisation at the expense of minor wellbeing.

The outcome of these investigations will carry implications well beyond California. Given the state's size and the tendency of other jurisdictions to follow its regulatory lead — particularly in technology policy — any consent decrees or settlements reached with the attorney general's office are likely to function as de facto national standards. The digital policy landscape is shifting rapidly, and platforms that have treated youth safety as a compliance checkbox rather than a product design principle are increasingly finding that approach unsustainable under sustained legal scrutiny.

For readers tracking the broader evolution of technology regulation and its intersection with safety, workforce, and governance issues, our coverage of how the UK's landmark AI safety legislation was passed into law offers important comparative context on how governments are reaching for statutory tools to manage technology risks that voluntary frameworks have failed to address.