Malaysia's communications regulator has taken down over 11,600 pieces of deepfake content since the beginning of 2024, marking an intensifying crackdown on artificial intelligence-generated misinformation that threatens the nation's digital landscape. Deputy Communications Minister Teo Nie Ching disclosed the removal figures during parliamentary questioning, underscoring the escalating battle against synthetic media that exploits advancing AI technology to deceive the public and potentially destabilise social discourse.
The scale of the deepfake problem facing Malaysia has expanded at an alarming pace. Complaints lodged with the Malaysian Communications and Multimedia Commission (MCMC) have exploded more than eightfold within just eighteen months, climbing from 917 in 2024 to 3,612 by early 2025, and reaching 7,967 as of mid-June this year. This geometric growth reflects both heightened public awareness of the threat and the genuine proliferation of synthetic content across messaging platforms and social networks, suggesting that digital literacy campaigns and regulatory action have prompted more reporting whilst the underlying problem intensifies.
To combat this surge in AI-facilitated deception, the government has enacted comprehensive new safeguards through the Risk Mitigation Code (RMC) embedded within the Online Safety Act 2025. These regulations mandate that licensed social media platforms implement sophisticated risk mitigation measures specifically designed to identify, monitor, and neutralise AI-generated content before it spreads virally. The framework represents a shift toward proactive platform responsibility rather than purely reactive enforcement, placing the burden on technology companies to invest in detection systems and content moderation infrastructure.
The MCMC has begun systematic engagement with major social media platform providers to assess their compliance with these new obligations. This oversight process involves evaluating whether platforms have established adequate safeguards, trained moderation teams, and deployed automated tools capable of identifying deepfakes and other synthetic content. The agency's compliance auditing approach signals that government scrutiny will be ongoing and potentially escalating, creating pressure on technology companies to demonstrate concrete progress in addressing the issue.
Beyond simply removing content, the MCMC has embedded itself into law enforcement investigations targeting AI misuse. The commission provides critical technical support to police and other agencies, including profiling information that identifies networks of coordinated inauthentic behaviour, digital forensic analysis that traces content origins, and metadata that links deepfakes to specific creators or distribution networks. This collaborative approach between communications regulators and traditional law enforcement represents an acknowledgment that tackling sophisticated digital deception requires expertise spanning both technical analysis and criminal investigation.
The regulatory framework also addresses the interconnected problem of fraudulent advertising enabled by deepfakes and fake accounts. Licensed platforms must now verify the identities of advertisers through official channels such as the Companies Commission of Malaysia before allowing advertisements to run. This verification requirement aims to prevent fraudsters from creating shell accounts and fake corporate entities to distribute scam advertisements, a vulnerability that has been extensively exploited in Southeast Asia to perpetrate financial fraud targeting vulnerable populations.
The consequences for platform non-compliance carry substantial financial weight. Companies failing to meet their obligations under the Risk Mitigation Code face potential criminal prosecution with maximum fines reaching RM1 million, supplemented by additional financial penalties of up to RM10 million. These cumulative penalties, which could total RM11 million per violation, represent genuinely punitive sanctions that should incentivise platform compliance, particularly for smaller operators or regional services with tighter profit margins.
For Malaysian readers and Southeast Asian observers, this regulatory tightening reflects broader regional trends toward stricter digital governance. Countries including Singapore, Thailand, and Indonesia have similarly moved to regulate AI content and social media conduct, though Malaysia's approach through the comprehensive Online Safety Act arguably provides clearer rules and more structured accountability mechanisms. The deepfake epidemic poses particular risks in the Malaysian context, where political polarisation, ethnic sensitivities, and religious diversity create conditions where synthetic media depicting controversial figures or communities could trigger widespread social tension.
The eightfold increase in complaints also suggests that public awareness campaigns about deepfakes and AI misuse have had some effect in encouraging reporting. However, the persistence of high complaint volumes indicates that awareness alone cannot solve the problem—structural interventions requiring platform cooperation and enforcement action remain essential. The government's willingness to impose substantial fines and pursue criminal prosecution demonstrates commitment to moving beyond awareness toward concrete deterrence.
Looking ahead, Malaysia's deepfake challenge will likely continue evolving as AI generation technology becomes more accessible and sophisticated. Current detection methods may become less effective as deepfake quality improves, requiring ongoing investment in more advanced identification systems. The MCMC and platform providers will need to maintain intensive research and development efforts to stay ahead of increasingly convincing synthetic content, whilst balancing security needs against legitimate uses of AI technology in entertainment, education, and creative industries.
The regulatory framework also carries implications for content creators and ordinary users. Individuals sharing or creating deepfakes, even for entertainment purposes, could face liability under the Online Safety Act provisions. This legal exposure may deter casual creation and sharing whilst potentially impacting legitimate creative expression, raising questions about where policymakers should draw lines between prohibiting genuinely harmful deception and permitting acceptable parody or artistic use of synthetic media.
Ultimately, Malaysia's experience demonstrates that effectively addressing AI-generated misinformation requires sustained coordination between government regulators, social media platforms, law enforcement agencies, and the public. The 11,600 removals represent tangible progress, yet the exponential growth in complaints suggests authorities are still chasing a rapidly expanding problem. Success will depend on whether platforms genuinely commit resources to detection systems, whether enforcement actions deter bad actors, and whether public literacy about deepfakes improves sufficiently to reduce vulnerability to deceptive content.
