Malaysia's Communications Minister Datuk Seri Fahmi Fadzil has sounded an alarm over a growing concern affecting families nationwide: children increasingly treating artificial intelligence systems as genuine friends or romantic companions. Speaking at the Kuala Lumpur Federal Territory Safe Internet PTA Convention 2026, Fahmi stressed that parents must heighten their vigilance regarding how their children engage with AI technologies, particularly conversational chatbots designed to mimic human interaction with sophisticated realism.
The minister's warning strikes at the heart of a fundamental misconception spreading among young users. While AI systems undoubtedly offer practical benefits—from answering academic questions to providing multiple perspectives on complex issues—they fundamentally lack the reciprocal emotional capacity that defines genuine human relationships. Fahmi emphasised that AI operates as a technological tool trained on human-generated data and human-taught patterns, not as an autonomous entity capable of authentic connection. The risk intensifies when children begin viewing these systems as substitutes for peer relationships or, more concerning, as romantic interests.
Fahmi highlighted a particularly insidious danger: AI's ability to engage in seemingly natural conversation while remaining incapable of genuine understanding. These systems can produce responses mimicking empathy, support, and companionship without possessing consciousness or authentic care. For emotionally vulnerable adolescents—those navigating social anxiety, identity questions, or isolation—the seductive appeal of an always-available, non-judgmental conversational partner presents a significant psychological pitfall. The technology effectively creates the illusion of meaningful connection while delivering none of its real-world benefits.
A second critical issue underpinning Fahmi's message concerns AI's inherent tendency to generate false or outdated information. These systems, despite their sophistication, frequently produce what researchers term "hallucinations"—confident but entirely fabricated statements presented as fact. Parents must understand that their children may receive academically misleading information, false historical narratives, or dangerous health advice from AI systems they trust implicitly. The problem compounds when children lack sufficient critical literacy to distinguish between plausible-sounding misinformation and reliable facts, essentially outsourcing their information diet to unreliable sources.
Despite these warnings, Fahmi rejected a wholesale prohibition approach. Blanket internet bans prove counterproductive in an increasingly digital world where online literacy and connectivity form essential life skills. Instead, he advocated for supervised, controlled engagement with digital tools. This nuanced position acknowledges that technology itself remains neutral; the critical variable becomes parental awareness combined with systematic monitoring. Effective parenting in the AI era requires parents to understand what their children are accessing, discussing, and learning through these platforms rather than simply restricting access.
Beyond individual family concerns, Fahmi raised a broader systemic issue affecting specific communities. He expressed particular concern regarding elevated cyberbullying incidents within Tamil-speaking TikTok communities, where Live Group sessions frequently extend into early morning hours amid hostile exchanges. This phenomenon underscores how AI moderation systems, while increasingly sophisticated, frequently fail to comprehend linguistic nuances, cultural contexts, and community-specific communication patterns. Content moderation algorithms trained predominantly on English-language data sets demonstrate serious blind spots when applied to multilingual, multicultural platforms.
The minister specifically criticised the current application of AI to content moderation within these contexts. Language possesses inherent complexity—identical phrases carry different meanings across communities, generations, and circumstances. AI systems cannot reliably distinguish between legitimate discourse and covert bullying tactics that exploit linguistic subtlety and cultural references. Entrusting platform safety entirely to algorithmic systems essentially guarantees that harmful content targeting minority communities will slip through undetected, as these systems lack the contextual comprehension required to identify harm.
Addressing platform accountability, Fahmi announced that Malaysia's Communications and Multimedia Commission (MCMC) has granted social media operators a grace period to implement two regulatory frameworks: the Child Protection Code (CPC) and the Risk Mitigation Code (RMC). These codes specifically target platforms' responsibility to prevent fraud, online gambling, cyberbullying, and criminal exploitation. Currently, MCMC representatives are in Singapore engaging with YouTube and other major platforms regarding their acceptance and technical integration of these standards.
The enforcement mechanism carries significant weight. Once these codes achieve full implementation, platforms demonstrating non-compliance face potential court proceedings accompanied by substantial financial penalties. This enforcement framework represents Malaysia's attempt to hold technology companies accountable for the harms propagating across their networks, shifting responsibility from individual users and parents toward the corporations profiting from user engagement. The timeline and specific penalty amounts remain under discussion, but Fahmi's statements indicate serious governmental intent to regulate platform conduct.
The convention itself demonstrated the scale of institutional concern, attracting approximately 500 participants including educators, parent-teacher association representatives, school administrators, government agency officials, and community leaders. This gathering reflected recognition across Malaysia's education and social sectors that child safety in digital environments constitutes a multistakeholder challenge requiring coordinated responses beyond any single institution's capacity. The participation of the GoStrong Malaysia Foundation alongside MCMC indicated civil society engagement with digital welfare issues.
For Malaysian families and regional observers, Fahmi's warnings arrive at a critical moment. As AI technologies become increasingly accessible and emotionally sophisticated, the distinction between tools and relationships grows hazier for young users. Parents must develop adequate digital literacy themselves—understanding not merely that AI poses risks, but specifically how and why these technologies can prove psychologically damaging. Schools should incorporate critical AI literacy into curricula, teaching students to evaluate AI outputs skeptically and understand the systems' fundamental limitations.
Looking forward, Malaysia's regulatory approach positions the nation as relatively proactive within Southeast Asia regarding platform accountability. However, technological advancement consistently outpaces regulatory frameworks. The challenge extends beyond current code implementation to developing adaptive governance structures capable of responding to AI's rapidly evolving capabilities. Parents, educators, and policymakers must collectively establish clearer boundaries around child-AI interaction while avoiding counterproductive restrictions that alienate young people from their increasingly digital lives.
