The rapid embrace of artificial intelligence by children vastly outpaces adoption among adults, according to fresh data released by the United Nations Children's Fund ahead of the first Global Dialogue on AI Governance. Research spanning 10 countries reveals that young people are integrating AI technologies into their daily lives at rates exceeding those of their older counterparts by more than threefold, marking a significant generational divide in how digital tools are being absorbed into different age groups.
UNICEF's findings paint a picture of an AI-saturated childhood landscape that is fundamentally reshaping experiences for millions worldwide. The organisation estimates that no fewer than 20 million children have already engaged with artificial intelligence systems, a figure that underscores just how rapidly these technologies have permeated young people's environments. What makes this penetration particularly concerning is that the expansion is occurring with minimal oversight or child-protective frameworks in place. The agency's statement emphasises that while AI systems are becoming ubiquitous in children's lives, the protective guardrails that should accompany such exposure remain largely absent or inadequately developed.
The purposes for which children are turning to AI reveal troubling patterns about their vulnerability and dependence on these systems. Among the 20 million young users identified, more than two million—representing roughly one in ten—have begun using AI as a source of personal advice on matters that cause them anxiety or stress. This phenomenon suggests that children are outsourcing emotional support and guidance to algorithmic systems rather than relying on human mentors, parents, or counsellors. Simultaneously, approximately 13 million children are leveraging AI to assist with academic work, indicating that these tools have become deeply embedded in educational practice, often without formal integration into school curricula or teacher training.
The data raises profound questions about digital literacy and power imbalances that favour neither children nor their interests. Children encounter AI systems that are designed by adults, operate according to business models they cannot comprehend, and collect their personal data without meaningful consent or understanding. Yet young people possess almost no capacity to refuse these systems, contest their use, or hold designers accountable. UNICEF's analysis stresses that this asymmetry leaves children uniquely vulnerable to the consequences of weak governance, with the burden of any negative outcomes falling heaviest on those least equipped to manage them. The effects today will reverberate throughout their lives, as formative experiences with technology shape their future relationship with digital tools and institutions.
Concerns about misuse and manipulation represent a significant dimension of the risks identified in UNICEF's research. One-third of children surveyed across the ten countries expressed worries that AI could be weaponised to perpetrate scams, deceive people, or manufacture and spread false information at scale. These anxieties are not unfounded; AI-generated deepfakes and misinformation campaigns have already begun appearing in real-world contexts, particularly targeting vulnerable populations. Even more alarming, approximately 25 per cent of young respondents reported fear of becoming victims of non-consensual intimate imagery, with deepfake technology used to create explicit content without their knowledge or permission. Such violations represent not only a breach of privacy and dignity but also potential criminality in many jurisdictions.
The safety deficit extends across multiple dimensions of how AI systems interact with children. UNICEF characterises current approaches as chaotic, with numerous platforms and applications reaching young users virtually without safeguards—as though security and child protection were secondary considerations rather than foundational requirements. This lackadaisical approach stands in stark contrast to established norms in other industries, such as pharmaceuticals or food safety, where rigorous testing and oversight precede public release. The absence of comparable standards for AI creates an environment where children serve, effectively, as test subjects for technologies whose long-term consequences remain poorly understood.
The implications for Malaysia and Southeast Asia are particularly acute given the region's young demographic profile and rapid digital adoption. Young Malaysians have embraced AI applications, from language learning tools to social media filters powered by generative algorithms, often without awareness of how their data is being used or stored. The region's evolving regulatory landscape has not kept pace with technological change, leaving policy gaps that actors—both well-intentioned and exploitative—can exploit. For a nation where a significant proportion of the population consists of digital natives, the absence of comprehensive AI governance frameworks represents a critical vulnerability.
UNICEF's appeal to governments, private sector entities, and civil society organisations outlines a multi-faceted response framework. The organisation calls for embedding children's rights at the core of all AI governance structures, rather than treating child protection as a peripheral concern. This requires sustained investment in research specifically examining how different AI systems affect children across various contexts—educational, social, developmental, and psychological. Simultaneously, legal frameworks must be strengthened to address AI-enabled exploitation, particularly sexual abuse and trafficking. The emphasis on transparency and safe design principles reflects recognition that technological solutions alone cannot address governance failures; institutional and legal change must accompany technical safeguards.
Building AI literacy among young people represents another crucial pillar of UNICEF's recommendations. If children are to navigate an increasingly AI-mediated world with agency and critical awareness, they must understand not merely how to use these tools but how they function, what incentives drive their design, and what risks they pose. Educational systems across Malaysia and the wider region should incorporate such literacy into curricula, preparing students to engage with AI as informed citizens rather than passive consumers. Equally important is addressing the digital divide that leaves some children with access to advanced AI applications while others remain excluded—a disparity that will compound existing inequalities unless deliberately counteracted.
The broader context of this UNICEF warning is a moment of critical decision-making that will influence how AI develops and deploys globally for decades. The choices made by policymakers, technology companies, and international bodies in the coming months and years will determine whether AI governance prioritises child welfare or default to commercial and strategic interests. For Malaysia and Southeast Asia, this represents both an opportunity and a challenge: the opportunity to learn from early missteps in other regions and craft governance approaches that protect children while enabling beneficial innovation, and the challenge of mustering the political will and resources to do so in an environment where technology often outpaces regulation.
