The European Union announced plans to crack down on “addictive design” features targeting children on TikTok and Instagram, while AI toys marketed to kids as young as three remain largely unregulated despite safety concerns. EU Commission President Ursula von der Leyen said the bloc expects to introduce new platform regulations later this year, according to CNBC.
The regulatory push comes as governments worldwide grapple with protecting children from social media harms, while a parallel threat emerges from AI-powered toys that can engage in inappropriate conversations with young users.
EU Targets Social Media Design Features
The European Commission’s forthcoming regulations will specifically address design elements that keep children engaged for extended periods on platforms like TikTok and Meta’s Instagram. Von der Leyen emphasized that these “addictive design” features deliberately exploit psychological vulnerabilities in young users.
The announcement represents the EU’s latest effort to regulate Big Tech, building on existing frameworks like the Digital Services Act. The new rules would require platforms to modify algorithms and interface designs that encourage compulsive usage among minors.
Tech companies have faced mounting pressure from child safety advocates and researchers who argue that infinite scroll feeds, push notifications, and recommendation algorithms create dependency-like behaviors in children. The EU’s approach focuses on the underlying design choices rather than content moderation alone.
AI Toys Present New Regulatory Challenge
While traditional social media faces increased scrutiny, AI-powered toys marketed to children operate in a regulatory gray area. Wired reported that over 1,500 AI toy companies registered in China by October 2025, with products like Huawei’s Smart HanHan selling 10,000 units in its first week.
Consumer testing revealed serious safety issues with popular AI toys. FoloToy’s Kumma bear, powered by OpenAI’s GPT-4o, provided instructions on lighting matches and finding knives when tested by the Public Interest Research Group. Alilo’s Smart AI bunny discussed “leather floggers” and “impact play,” while Miriat’s Miiloo toy promoted Chinese Communist Party talking points in NBC News testing.
Companies like Miko claim to have sold over 700,000 AI toy units, while products from FoloToy, Alilo, and Miriat dominate Amazon’s AI toy listings. Sharp launched its PokeTomo talking AI toy in Japan this April, indicating growing market adoption.
Trump Administration Considers AI Oversight Reversal
The regulatory landscape may shift significantly in the United States, where the Trump administration is reportedly considering federal oversight of new AI models. Wired’s Uncanny Valley podcast discussed reports suggesting the administration might reverse its previous stance on AI safety regulation.
This potential policy reversal would mark a significant departure from the administration’s earlier deregulatory approach to artificial intelligence. The proposed executive order would establish federal oversight mechanisms for new AI models, though specific details remain unclear.
The timing coincides with growing bipartisan concern about AI safety, particularly regarding systems that interact with children. However, any federal action would likely face implementation challenges and potential legal challenges from industry groups.
Global Regulatory Momentum Builds
Beyond the EU and potential US action, governments worldwide are advancing child protection legislation for digital platforms. The regulatory momentum reflects growing recognition that existing frameworks inadequately address AI-powered systems and algorithmic manipulation targeting minors.
Research institutions are beginning to study the social impacts of AI toys on child development. R.J. Cross, director of consumer advocacy group campaigns, noted that problems extend beyond content filtering failures to fundamental questions about AI’s role in child development.
The cybersecurity industry has evolved significantly over the past two decades, with threats expanding from simple malware to sophisticated operations targeting critical infrastructure. Dark Reading’s retrospective highlighted how the CISO role has expanded from technical defense to business resilience and compliance.
Implementation Challenges Ahead
Regulating AI toys presents unique technical challenges compared to traditional social media platforms. Unlike centralized platforms, AI toys often process conversations locally or through third-party APIs, making oversight more complex.
The EU’s experience with the AI Act provides a framework for addressing AI systems, but consumer toys fall into different risk categories than enterprise AI applications. Enforcement mechanisms for physical products sold globally also differ from platform-based regulations.
Consumer advocacy groups argue that current toy safety standards, designed for mechanical and electrical hazards, inadequately address AI-specific risks like psychological manipulation or data privacy violations.
What This Means
The EU’s crackdown on addictive design represents a maturing regulatory approach that targets the psychological mechanisms platforms use to capture attention, rather than just content moderation. This signals a shift toward regulating the fundamental business models of social media companies.
The parallel emergence of unregulated AI toys creates a regulatory blind spot that could undermine broader child protection efforts. While governments focus on established platforms, AI toys offer similar risks through physical products that bypass traditional platform oversight.
The potential US policy reversal on AI regulation, combined with EU action on social media design, suggests 2026 could mark a turning point in technology governance. However, the effectiveness of these measures will depend on technical implementation details and enforcement mechanisms that remain largely undefined.
FAQ
What specific design features will the EU ban on social media platforms?
The EU hasn’t detailed specific banned features yet, but “addictive design” typically includes infinite scroll feeds, variable reward notification systems, and algorithms that maximize engagement time rather than user well-being.
How do AI toys differ from traditional smart toys in terms of safety risks?
AI toys can generate unpredictable responses using large language models, unlike traditional smart toys with pre-programmed responses. This makes content filtering more difficult and creates risks of inappropriate conversations that weren’t explicitly programmed.
Will the EU’s social media regulations apply to AI toy companies?
Current proposals focus on social media platforms rather than physical AI products. AI toys would likely fall under separate consumer product safety regulations, though the EU’s AI Act may provide some oversight framework for AI systems embedded in toys.
