Brussels Presses TikTok to Redesign Features for Child Safety

EU regulators are escalating pressure on TikTok, arguing the app’s design nudges children and teens into compulsive use. In a preliminary decision tied to the Digital Services Act (DSA), the European Commission says features like endless scrolling, autoplay and push notifications must be changed, with stronger tools to encourage breaks—especially at night.

What the EU alleges TikTok is doing wrong

The Commission’s preliminary findings focus on what it describes as “addictive” or “rewarding” product mechanics. According to the investigation, TikTok has not adequately evaluated how these mechanics could harm users physically and psychologically, including minors and vulnerable adults.

The EU highlights several specific design patterns:

• Endless scroll that continuously serves new content
• Autoplay that reduces friction between videos
• Push notifications that re-engage users and pull them back into the app
• A “rewarding” feed dynamic that can put the brain on “autopilot,” encouraging prolonged sessions

Regulators argue these patterns can reduce self-control and contribute to compulsive behavior. The Commission also flags age verification as a persistent weakness, warning that inadequate controls can expose minors to inappropriate content.

The DSA context: risk assessments, not just content moderation

This case matters because it shows how the DSA is being applied beyond illegal content takedowns. The law also requires large online platforms to assess and mitigate “systemic risks,” including risks to minors’ mental health and well-being.

In practice, that means regulators can scrutinize product design choices—recommendation systems, engagement loops, and interface defaults—when they plausibly drive harmful outcomes. The Commission’s critique that TikTok did not “adequately” evaluate risks points to a core DSA obligation: platforms must perform rigorous risk assessments and demonstrate that mitigations are effective, not merely present.

The EU also argues TikTok’s existing mitigations fall short. Tools such as screen-time limits and parental controls are described as too easy to dismiss, implying that “safety features” must be designed with real-world user behavior in mind, especially for teens.

TikTok’s response and what happens next

TikTok disputes the Commission’s position. A company spokesperson said the preliminary conclusions present a categorically incorrect and unfounded picture of the platform and that TikTok will contest the findings using all available means.

Procedurally, TikTok now has the right to review the evidence and defend itself. If the Commission ultimately finds a violation, penalties can reach up to 6% of global annual turnover—an enforcement lever large enough to influence product roadmaps.

TikTok reported more than 200 million users in Europe, making the region strategically important for its advertising business and creator ecosystem. That scale also increases regulatory expectations under the DSA, which targets “very large online platforms” with heightened duties.

Why this is a pivotal moment for AI-driven feeds

Although the Commission’s language focuses on “design,” the underlying engine is AI. TikTok’s core experience is powered by recommender systems that optimize for engagement signals—watch time, replays, likes, shares, follows, and session length. Modern ranking models can learn highly individualized preferences quickly, especially when fed continuous behavioral data.

This creates a broader industry dilemma: the same machine-learning techniques that deliver relevance and discovery can also intensify “stickiness.” When combined with autoplay, infinite scroll, and notifications, the system reduces stopping cues—an issue researchers often describe as a shift from user choice to algorithmic momentum.

The EU’s approach signals that regulators may increasingly treat engagement-optimized recommender systems as a safety and compliance issue, not merely a product feature. That could push platforms toward:

• Stronger default limits for minors (night-time restrictions, break prompts)
• More robust age assurance, potentially including privacy-preserving verification methods
• Greater transparency and auditing around recommender goals and outcomes
• Design changes that add friction (confirmation prompts, feed pauses, reduced notification intensity)

The broader child-safety landscape in Europe

The Commission’s move also reflects political urgency around youth well-being. In Sweden, for example, social media use among children and teens is nearly universal: 97% of 8–19-year-olds use social media and 86% use it daily, according to the Swedish Internet Foundation’s “Children and the Internet 2025” report. TikTok is among the top platforms alongside YouTube, Snapchat, Instagram and Roblox.

For the AI industry, this case is a reminder that “responsible AI” is not confined to model bias or privacy. It increasingly includes how AI systems shape attention, habits, and mental health—especially for minors. If the EU finalizes its preliminary view, TikTok’s outcome could become a template for how Europe regulates engagement-centric design across AI-driven consumer platforms.