The United Nations Children's Fund has raised urgent concerns about the pace at which young people are embracing artificial intelligence, revealing that children are adopting these technologies at more than three times the rate of their adult counterparts. Speaking ahead of the first Global Dialogue on AI Governance in Hamilton, Canada, UNICEF presented findings from research spanning 10 countries that paint a picture of rapid, largely unregulated AI integration into childhood experiences across the globe. The organisation emphasised that while these technologies are reshaping how children learn, play, and interact with the world, the infrastructure to protect them remains dangerously underdeveloped.
UNICEF's research indicates that at least 20 million children globally have already engaged with AI systems in some form. What distinguishes this figure is not merely its scale, but the nature of the interactions. More than two million children—representing roughly one in 10 of those using AI—have turned to these systems for guidance on personal matters that concern them, effectively treating algorithms as confidants for worries and anxieties. This trend underscores a fundamental shift in how the youngest generation seeks support and counsel, one that occurs largely outside parental oversight or institutional safeguards. For educators and parents across Southeast Asia and beyond, the implications are profound: children are forming relationships with AI in ways previous generations never experienced.
Within the educational sphere specifically, approximately 13 million of the surveyed children reported using AI to assist with their studies and homework. This application mirrors patterns seen in developed economies where AI tutoring systems and content generators have proliferated rapidly. Yet the accessibility and deployment of such tools often outpace any meaningful evaluation of their pedagogical value or potential to reinforce existing inequalities in access to quality education. The rush to harness AI's perceived educational benefits has occurred with minimal consideration of how different children, particularly those in lower-income regions, experience these technologies differently—a concern especially relevant for Malaysia and other developing nations in the region.
The organisation's statement articulated a critical asymmetry at the heart of this digital transformation. Children face unprecedented exposure to AI systems without commensurate power to understand, question, or opt out of them. This powerlessness extends across multiple dimensions: they rarely comprehend how these systems are constructed, what commercial interests drive them, or how their own personal data fuels their operations. UNICEF stressed that children will inevitably encounter the most severe consequences of weak AI governance frameworks, yet possess the least agency in shaping the rules that govern these technologies. This generational vulnerability distinguishes AI adoption among youth from previous technology transitions.
The safety concerns identified in UNICEF's research are particularly alarming. Among the 10 countries surveyed, one third of children expressed anxiety about AI being weaponised for fraud, deception, and the dissemination of false information. These worries reflect genuine threats already materialising in digital spaces worldwide. However, an even more disturbing finding concerns deepfake technology: a quarter of the children surveyed feared the possibility of their images or videos being manipulated into sexually explicit material without consent. This specific fear points to an emerging abuse vector that combines AI's technological capabilities with the sexual exploitation of minors—a crime that exists largely outside traditional regulatory frameworks. The psychological toll of such vulnerabilities on young people, particularly in societies with robust online cultures, deserves serious attention from policymakers and technology platforms alike.
UNICEF's characterisation of the current landscape as one where numerous AI systems reach children with "no guardrails" and safety treated as "seemingly an afterthought" reflects the tension between rapid technological deployment and responsible governance. Across the globe, technology companies have prioritised speed to market and user growth over implementing meaningful protections for vulnerable populations. This pattern is especially evident in the Global South, where regulatory capacity is often limited and market pressures from international tech giants dominate. The absence of safeguards is not accidental; it reflects deliberate prioritisation of commercial imperatives over child welfare.
The fund has issued a comprehensive agenda for reform that demands action from multiple stakeholders. Governments must prioritise research into AI's actual and potential harms affecting children, moving beyond theoretical risks to understand real-world impacts in their specific contexts. Legislative frameworks need strengthening, particularly regarding crimes enabled by AI such as sexual exploitation and abuse. Technology companies must redesign their systems with child safety and transparency as foundational principles rather than afterthoughts, ensuring parents and educators can understand how these tools function. Educational initiatives must build AI literacy among young people themselves, enabling them to navigate these systems with greater awareness. Simultaneously, closing the digital divide remains essential—ensuring that access to AI's educational and beneficial applications is not limited to wealthy populations.
The timing of UNICEF's call for action coincides with a critical juncture in global AI governance. International bodies, national governments, and industry stakeholders are currently engaged in discussions that will shape regulatory approaches for years to come. The choices made during this period—regarding safety standards, data protection, algorithmic transparency, and accountability mechanisms—will have lasting consequences for children worldwide. Malaysia and other Southeast Asian nations have an opportunity to advocate for child-centric governance frameworks that prioritise protection without unnecessarily limiting beneficial innovation. This requires moving beyond treating child safety as an optional add-on to becoming a central organising principle of AI development and deployment.
For Malaysian society specifically, these findings should prompt reflection on several fronts. The nation's growing digital sophistication and high youth internet penetration suggest that children here are likely adopting AI at rates consistent with or exceeding global averages. Yet regulatory frameworks governing AI remain nascent, and public awareness of these technologies' risks remains limited. Parents, educators, and policymakers across the region lack comprehensive guidance on supporting children's healthy engagement with AI while mitigating identified harms. Building this capacity requires urgent investment in research, education, and policy development tailored to Southeast Asian contexts.
UNICEF's framing of this moment as "decisive" reflects the genuine stakes involved. Artificial intelligence is not a hypothetical future technology for children but a present-day reality shaping their learning, safety, and development. The decisions made about how these systems are designed, deployed, and governed will reverberate across generations. Allowing current trends to continue—where children adopt AI at accelerating rates with minimal protection—represents an abdication of collective responsibility toward vulnerable populations. Conversely, embracing comprehensive child-centred governance frameworks offers an opportunity to harness AI's genuine benefits while building in meaningful protections. For Malaysia and the broader region, responding thoughtfully to UNICEF's call requires moving beyond treating AI regulation as a technical matter for experts alone, toward recognising it as a fundamental question of child welfare and social justice.
