Malaysia’s government may reassess social media licensing rules for X after 17 complaints about Grok AI misuse, with MCMC considering legal action.
PETALING JAYA: The government may reassess the eight-million-user threshold that currently exempts social media platform X from licensing requirements in Malaysia, following repeated misuse of its artificial intelligence application, Grok.
Communications Minister Datuk Fahmi Fadzil said online harm does not cease simply because a platform has fewer than eight million users.
Speaking in the Dewan Rakyat yesterday, he said the Malaysian Communications and Multimedia Commission (MCMC) has recorded 17 complaints linked to Grok, comprising six formal complaints, two police reports, eight complaints flagged via social media and one initial First Information Report opened by MCMC.
Fahmi said MCMC is examining the hybrid nature of platform X and Grok to ensure harmful content can be addressed comprehensively. During a meeting with X representatives on Wednesday, the company confirmed that tighter controls had been imposed on Grok’s image and video features, preventing misuse to generate or edit pornographic, sexual or incest-related content.
“I have requested formal confirmation of the measures taken so MCMC can assess their effectiveness and determine appropriate regulatory action,” he said, adding that the steps are necessary to protect Malaysians, especially children and vulnerable groups, and align with Malaysian laws, the Rukun Negara and international practices.
MCMC temporarily blocked access to Grok on Jan 11 after repeated misuse. Fahmi said the restriction is not permanent and may be lifted if safeguards are verified through testing. MCMC is also considering potential legal action against X for non-compliance with online safety regulations.
Meanwhile, Universiti Malaysia Kelantan Institute for Artificial Intelligence and Big Data research fellow Dr Siti Salina Saidin cautioned that risks of AI misuse remain despite new restrictions.
She said experienced users can bypass safeguards through prompt manipulation, tool chaining or open-source models without built-in guardrails.
She defended the government’s move, saying generative AI is not morally neutral and that early intervention is a form of risk containment, not a rejection of innovation.
