The adoption of artificial intelligence among children has reached concerning levels, with the United Nations Children's Fund reporting that the world's young people are embracing the technology at rates more than three times faster than their adult counterparts. This disparity, revealed through research spanning 10 countries, underscores a fundamental mismatch between the rapid proliferation of AI systems accessible to children and the regulatory safeguards designed to protect them. The findings, presented ahead of the inaugural Global Dialogue on AI Governance, paint a picture of a digital landscape that has fundamentally transformed childhood, creating opportunities alongside unprecedented vulnerabilities that policymakers worldwide have been slow to address.
The scope of AI penetration into children's lives is substantial. UNICEF's research indicates that at minimum 20 million children across the surveyed nations have engaged with artificial intelligence technologies, whether through educational applications, entertainment platforms, or digital assistants. More alarming is the statistic that approximately one in ten children—more than two million globally—have begun relying on AI systems to provide guidance on matters causing them anxiety or emotional distress. This represents a profound shift in how young people seek support and counsel, circumventing traditional sources of advice from parents, teachers, and counsellors, and placing vulnerable young minds in the hands of algorithms whose transparency and accountability remain largely opaque.
The educational applications of AI have also gained substantial traction among children. An estimated 13 million young users leverage artificial intelligence to assist with their schoolwork and learning pursuits, reflecting the growing integration of these technologies into both formal and informal education systems. While such applications can offer personalised learning pathways and democratise access to educational resources, they simultaneously raise questions about data privacy, algorithmic bias in educational assessment, and the appropriate role of machine learning in shaping the developmental experiences of young learners. Schools and parents across Malaysia and the broader Southeast Asian region have begun grappling with these implications as AI-powered tutoring platforms and homework assistance tools proliferate.
The governance vacuum surrounding children's exposure to these technologies represents perhaps the most pressing concern identified by UNICEF. The agency emphasises that children face extensive exposure to AI systems—encompassing the design philosophies embedded within them, the commercial incentives driving their development, and the data harvesting mechanisms supporting their operations—yet possess minimal agency in avoiding or challenging these technologies. This asymmetry of power and knowledge places children in a fundamentally precarious position, unable to fully comprehend or consent to the terms under which they interact with increasingly sophisticated digital systems.
Specific harms identified in the research warrant immediate attention. One-third of children surveyed expressed concerns regarding AI's potential misuse for deception, fraud, and the dissemination of false information, highlighting their awareness that AI technologies can be weaponised to manipulate and mislead. Even more troubling, one-quarter of young respondents reported anxiety about artificial intelligence being weaponised to generate and distribute sexually exploitative deepfakes bearing their likenesses. These fears are not merely theoretical; documented cases of such abuse have already emerged globally, demonstrating that children's apprehensions reflect genuine, evolving threats rather than abstract anxieties.
The absence of appropriate protective measures compounds these dangers. UNICEF's characterisation of current conditions—where numerous AI systems reach children effectively unguarded, with safety considerations treated as secondary—reflects a systemic failure to prioritise child welfare in the digital economy. Technology companies have frequently adopted a reactive rather than proactive approach to child safety, implementing protective measures only after widespread harm has been documented and public outcry has intensified. This pattern extends across jurisdiction after jurisdiction, with regulators typically lagging behind technological innovation and children bearing the human costs of this enforcement gap.
For Malaysia and Southeast Asia specifically, these findings carry particular resonance. The region has experienced rapid digital adoption, with smartphone penetration and internet access expanding significantly over the past decade. Young people in Malaysia, Indonesia, Thailand, and neighbouring nations are often early adopters of global digital platforms and services, including AI-powered applications. However, regulatory frameworks governing AI safety and child protection remain underdeveloped across most Southeast Asian jurisdictions, creating a context where children may face heightened exposure to poorly-governed AI systems while benefiting from comparatively fewer protective regulations than their counterparts in more heavily-regulated markets.
UNICEF has articulated a comprehensive agenda for responsible AI governance centred on child rights. The organisation calls upon governments to strengthen legal protections specifically targeting AI-enabled sexual exploitation and abuse, recognising that existing legislation often proves inadequate for addressing harms enabled by emerging technologies. Simultaneously, the agency advocates for corporate accountability, urging technology developers to embed child-safety principles into the architecture of systems from inception rather than treating protection as an afterthought. Transparent design practices, the agency argues, must become standard, allowing independent researchers and advocates to scrutinise AI systems for potential harms before they are deployed at scale.
Developing digital literacy among children represents another critical component of UNICEF's recommended approach. As young people increasingly navigate AI-powered environments, equipping them with sophisticated understanding of how these technologies function, what data they collect, and how their personal information is commercialised becomes essential. Such literacy programmes should extend beyond basic technological competency to encompass critical evaluation of algorithmic outputs, recognition of manipulative design patterns, and understanding of individual rights in digital contexts. Schools across Malaysia would benefit from integrating such curricula into existing educational frameworks.
Addressing the digital divide remains integral to equitable AI governance affecting children. While the most visible concerns often focus on wealthy nations where children have abundant access to advanced AI systems, the converse problem—insufficient access to beneficial AI-powered educational and developmental resources—affects children in lower-income settings. Genuine child-centred AI governance must therefore ensure that protective measures do not inadvertently deny disadvantaged populations access to technologies that could enhance learning opportunities and economic mobility. This balancing act presents particular challenges for developing nations in Southeast Asia seeking to harness AI's potential while protecting vulnerable youth populations.
The urgency of UNICEF's message cannot be overstated. The organisation emphasises that decisions made regarding AI governance in the coming years will reverberate throughout children's lives for decades, shaping their safety, privacy, well-being, and access to opportunity. The current moment represents a critical juncture where establishing robust protective frameworks remains possible, but windows of opportunity may narrow as these technologies become more deeply embedded in social, educational, and economic structures. Policymakers, technology companies, educators, and parents across Malaysia and the region must act decisively to ensure that artificial intelligence serves children's genuine interests rather than commercial imperatives alone.
