Malaysia Mandates Social Media Compliance with Child Protection Codes

Malaysian social media platform operators will be required to adhere to the Child Protection Code (CPC) and the Risk Mitigation Code (RMC) within the next few months. Minister of Communications, Datuk Seri Fahmi Fadzil, announced that failure to comply with these codes will result in legal action.

Fahmi explained that the grace period is necessary to allow platform operators sufficient time to fully understand the codes and make the required system adjustments. "Previously, they asked for a little time. First, to fully understand the codes in question and second, to operationalize them," he said. "So, we are giving them a little time, perhaps a period of a few months, to implement them." Once fully enforced, platforms that do not comply could face significant legal penalties and substantial fines.

“

Previously, they asked for a little time. First, to fully understand the codes in question and second, to operationalize them. So, we are giving them a little time, perhaps a period of a few months, to implement them.

The Malaysian Communications and Multimedia Commission (MCMC) is currently in discussions with several platform operators, including YouTube, in Singapore regarding the acceptance and implementation of the CPC and RMC. The minister also expressed concern over the increasing reliance of social media platforms on artificial intelligence (AI) for content moderation, particularly when it involves specific language nuances. He noted that AI's limitations in understanding the context of language could lead to criminal content or cyberbullying escaping moderation processes.

"The issue of language cannot be handed over entirely to AI. That tool is not capable of fully understanding the context of certain sentence usage, especially those that victimize (bully) others," Fahmi stated. He specifically pointed to TikTok's approach, noting that the platform has reduced its human moderators, focusing more on AI-driven moderation. This shift raises concerns about the effectiveness of content filtering, especially for subtle forms of online harm.

“

The issue of language cannot be handed over entirely to AI. That tool is not capable of fully understanding the context of certain sentence usage, especially those that victimize (bully) others.