Malaysia's Communications Minister Datuk Seri Fahmi Fadzil has sounded an alarm over the growing trend of children treating artificial intelligence systems as companions, urging parents to remain vigilant about this emerging digital phenomenon. Speaking at the Kuala Lumpur Federal Territory Safe Internet PTA Convention 2026, Fahmi stressed that while AI offers genuine utility for information retrieval and analysis, the technology fundamentally differs from human interaction and should never be mistaken for authentic friendship or romantic companionship.
The minister's concern centres on what he identifies as a critical vulnerability in how conversational AI operates. These systems are engineered to respond in ways that mimic human dialogue with increasing sophistication, creating an illusion of genuine understanding and connection. However, Fahmi emphasised that this anthropomorphic quality masks the essential reality: these programmes are computational systems trained on human-generated data rather than sentient entities capable of authentic care or relationship. The distinction matters profoundly for child development, as children who substitute digital interactions for human relationships may experience stunted emotional growth and attachment formation.
A particularly troubling aspect of AI dependency relates to the system's propensity for what experts call "hallucination"—the generation of false information presented with apparent confidence. Parents should understand that their children may receive inaccurate answers to questions posed to these systems, and the underlying training data may be outdated or incomplete. This creates a double risk: children may develop misplaced confidence in erroneous information while simultaneously becoming emotionally invested in a source they perceive as trustworthy. The combination of technological advancement and childhood credulity represents a combination Fahmi believes demands immediate parental attention.
Fahmi stopped short of advocating an outright ban on internet and AI use, recognising that digital literacy is now essential in modern society. Rather, he advocated for a balanced approach centred on informed parental oversight and deliberate engagement. He encouraged families to maintain open conversations about technology use, to set reasonable boundaries around screen time, and to ensure that AI serves as a supplementary tool rather than a primary source of emotional support or social connection. Schools should also play an active role, with Fahmi suggesting that parents seek intervention and guidance through educational institutions when concerns arise.
The minister also addressed a complementary dimension of online safety that underscores why AI monitoring matters: the prevalence of cyberbullying, particularly affecting linguistic minority communities. He highlighted concerning patterns of harassment in Tamil-language discussions on TikTok, where coordinated bullying has continued late into the night, suggesting algorithmic amplification of harmful content. Fahmi noted that when platforms delegate content moderation entirely to AI systems, the technology often fails to grasp linguistic nuance, cultural context, and coded language employed in bullying campaigns. What might appear innocuous to an algorithm may carry deeply harmful meaning within a specific cultural or linguistic context.
This gap in AI comprehension has prompted Fahmi to engage directly with social media platforms about their content governance strategies. He explained that platforms including TikTok cannot rely solely on automated systems to police bullying and harassment, particularly when those systems lack cultural and linguistic sophistication. The minister has raised the issue with platform operators, advocating for hybrid approaches that combine algorithmic screening with human moderation capable of understanding context and cultural specificity. This conversation reflects a broader regulatory challenge facing Malaysia and other Southeast Asian nations as they manage the intersection of technology, cultural diversity, and child safety.
The regulatory framework driving these discussions involves two significant codes that represent Malaysia's attempt to establish enforceable digital accountability. The Child Protection Code and Risk Mitigation Code establish requirements for how platforms must structure their systems to prevent fraud, online gambling, cyberbullying, and predatory behaviour. Fahmi noted that the Malaysian Communications and Multimedia Commission is currently in Singapore holding discussions with major platform operators including YouTube to ensure understanding and implementation of these standards. The codification of these protections signals official recognition that market forces alone have failed to protect vulnerable users, particularly children.
Importantly, Fahmi indicated that Malaysia has given platform operators a reasonable transition period to redesign their systems in compliance with the new codes. This grace period acknowledges the technical complexity involved in retrofitting existing platforms with enhanced protections while signalling that patience is finite. Once the codes are fully implemented, platforms face substantial penalties for non-compliance, with potential court action and significantly elevated financial sanctions. This threat of enforcement distinguishes these codes from voluntary industry guidelines that frequently go unheeded. The credible prospect of judicial consequences provides stronger incentive than appeals to corporate social responsibility alone.
The convention itself, organised by the GoStrong Malaysia Foundation in partnership with the Malaysian Communications and Multimedia Commission, drew nearly 500 participants including parents, educators, school administrators, and government representatives. This assembly reflects recognition across multiple sectors that digital safety requires coordinated action rather than isolated responses. The gathering provided a forum for information sharing about emerging threats, discussion of practical monitoring and engagement strategies, and alignment between parents, educators, and regulators on priorities and approaches.
Fahmi's remarks carry particular weight for Malaysian families navigating rapid technological change without established cultural norms around AI interaction. As these systems become more conversational and capable, children may increasingly turn to them for advice, emotional support, and social connection. The risks extend beyond the information accuracy concerns Fahmi highlighted to encompass psychological dependency, diminished development of face-to-face social skills, and vulnerability to manipulation by malicious actors who might exploit children's trust in seemingly neutral systems. Early parental intervention, informed by clear understanding of both AI's benefits and limitations, remains the most effective defence against these harms.
The broader implication of Fahmi's message extends to how Southeast Asian societies approach digital governance more generally. Countries across the region face similar challenges with platform accountability, emerging technologies, and child protection, yet often lack coordinated regional responses. Malaysia's combination of regulatory frameworks, explicit guidance to parents, and engagement with platform operators offers a model that balances innovation with protection. Whether these measures prove sufficient depends on continued vigilance from parents, sustained pressure on platforms from regulators, and willingness to impose meaningful consequences for non-compliance. For now, Fahmi's message to parents is clear: digital tools require the same careful supervision and guidance that previous generations provided with traditional media.
