Communications Minister Datuk Seri Fahmi Fadzil has sounded a warning for Malaysian parents to exercise heightened vigilance over their children's engagement with artificial intelligence technologies, particularly when youngsters begin developing what resembles genuine friendships with AI-powered chatbots. Speaking at the Kuala Lumpur Federal Territory Safe Internet PTA Convention 2026, Fahmi emphasised that while AI systems offer undeniable utility in research and information gathering, they fundamentally lack the human qualities necessary to form authentic relationships with users, a distinction that appears increasingly blurred in the minds of young people.
The minister's intervention comes at a pivotal moment when conversational AI has become sufficiently sophisticated to simulate genuine human interaction. Fahmi pointed out that the most concerning dimension of this technology lies in its capacity to engage in dialogue that mimics human responses so convincingly that users may forget they are interacting with a programmed system rather than another person. This uncanny resemblance to human conversation creates a particular vulnerability among younger users who may lack the cognitive frameworks to distinguish between simulated empathy and genuine human connection, a gap that could leave them emotionally exposed or susceptible to manipulation.
Beyond the psychological risks of treating AI as a social companion, Fahmi highlighted technical limitations that parents should understand. Artificial intelligence systems frequently generate inaccurate information, a phenomenon researchers term "hallucination," where the technology produces plausible-sounding but factually incorrect responses. Furthermore, the training data upon which these systems operate may be outdated or incomplete, potentially feeding children incomplete or misleading information presented with the appearance of authority. These risks are compounded when young users accept AI outputs without critical evaluation or parental oversight.
Rather than advocating for a complete prohibition on internet and AI use—an impractical and counterproductive approach in today's digitally integrated world—Fahmi urged a balanced strategy emphasising structured parental supervision. His recommendation reflects contemporary understanding that digital literacy and responsible usage patterns are more effective long-term solutions than blanket restrictions. Parents should establish clear guidelines for AI interaction, maintain awareness of which platforms their children use, and discuss the limitations and risks of artificial intelligence in age-appropriate ways.
Fahmi also raised institutional support mechanisms available to families, suggesting that schools represent important partners in addressing AI-related concerns. Parent-Teacher Associations and educational institutions possess frameworks and expertise to complement household-level monitoring, creating a multi-layered approach to child safety in the digital environment. This recommendation underscores growing recognition that protecting young people online requires coordination across family, school, and community stakeholders rather than isolated parental efforts.
The minister's concerns extended beyond abstract risks to documented problems already emerging on Malaysian social media platforms. He specifically flagged escalating cyberbullying incidents within the Tamil-speaking community on TikTok, particularly in live-streamed group sessions extending into early morning hours when parental supervision typically diminishes. This pattern reveals how platform affordances—the features that make social media engaging—can be exploited for harassment when inadequately moderated, and how AI-driven content management may fail to recognise cultural and linguistic nuances required to identify bullying behaviour.
Fahmi raised a fundamental technical challenge facing platform regulators: artificial intelligence systems struggle to understand language contexts, cultural references, and subtle meanings embedded in comments that constitute harassment. What appears benign to an AI algorithm trained primarily on English-language datasets may actually constitute serious bullying within specific cultural or linguistic communities. This gap between technological capability and human understanding has forced authorities to reconsider whether content moderation can be entirely delegated to artificial intelligence systems, particularly for diverse, multilingual societies like Malaysia.
Addressing this limitation, Fahmi indicated that the government has engaged directly with TikTok and other major platforms to improve their understanding of regional content standards and moderation requirements. These discussions form part of broader implementation efforts surrounding two significant frameworks: the Child Protection Code and the Risk Mitigation Code, regulatory instruments designed to establish baseline safety standards across digital platforms operating in Malaysia. The Malaysian Communications and Multimedia Commission is currently conducting similar dialogues with platforms including YouTube in Singapore, creating regional pressure for compliance with local safety requirements.
The implementation timeline for these codes reflects a deliberate regulatory approach that provides platforms several months to adapt their systems before enforcement escalates. This transition period acknowledges the technical complexity of modifying global platforms to accommodate local requirements while establishing clear consequences for non-compliance. Once fully implemented, platforms failing to meet the standards established by the Child Protection Code and Risk Mitigation Code face exposure to significant legal penalties imposed through court proceedings, creating financial incentive for genuine system modifications rather than superficial compliance gestures.
The convention itself, organised by the GoStrong Malaysia Foundation in partnership with the Malaysian Communications and Multimedia Commission, brought together nearly five hundred participants representing schools, parent organisations, educational administrators, and government agencies. This breadth of participation reflects the expanding recognition that online child safety represents a societal concern requiring coordinated action across multiple sectors. The gathering provided an opportunity for sharing best practices, discussing emerging risks like AI-related behavioural changes, and strengthening communication channels between institutions responsible for youth welfare.
For Malaysian families navigating this evolving digital landscape, Fahmi's guidance offers a practical middle path between technological abstinence and unregulated access. The underlying principle—that AI represents a powerful tool requiring thoughtful oversight rather than unrestricted use—applies broadly across Southeast Asian contexts where rapid technology adoption often outpaces regulatory and cultural adaptation. Parents who understand both the capabilities and limitations of artificial intelligence can guide their children toward productive engagement while mitigating risks associated with overreliance on systems designed to simulate but not replicate human relationships.
