The Malaysian government has escalated its fight against artificial intelligence-generated deception, removing over 11,600 deepfake videos and manipulated media from social platforms following formal takedown requests by the Malaysian Communications and Multimedia Commission. Deputy Communications Minister Teo Nie Ching disclosed the figure during parliamentary questioning, underscoring the scale of the nation's effort to combat synthetic content that threatens information integrity and public trust across Southeast Asia's digital landscape.
The trajectory of complaints registered with authorities paints a troubling picture of deepfake proliferation. The number of reports grew dramatically from 917 cases during 2024 to 3,612 in 2025, with the rate accelerating further to 7,967 complaints by mid-June this year. This eightfold increase in reporting reflects both greater public awareness of the threat and the genuinely alarming spread of sophisticated AI tools capable of creating convincing but false video and audio content. For Malaysia, a nation where social media penetration exceeds regional averages and political discourse frequently plays out online, this escalation poses particular risks to electoral integrity and social cohesion.
The government has responded by embedding new safeguards into the regulatory framework through the Risk Mitigation Code, a central provision within the Online Safety Act 2025 that became law this year. Under these requirements, all licensed social media platform operators must now implement comprehensive risk mitigation measures specifically designed to address artificial intelligence misuse. These obligations extend beyond reactive removal of flagged content to encompass proactive systems, content verification protocols, and technical infrastructure designed to detect and suppress synthetic media before it spreads widely. The approach reflects recognition that traditional moderation alone cannot contain the velocity and sophistication of AI-generated falsehoods.
Implementation oversight remains a critical element of Malaysia's strategy. The MCMC has initiated systematic engagement with platform operators holding broadcast licenses to assess their compliance with the new Risk Mitigation Code obligations. These compliance assessments move beyond theoretical commitments to evaluate whether platforms have actually deployed the technological and human resources necessary to identify and remove deepfake material at scale. The commission's role as a liaison between regulators and private operators addresses a fundamental challenge in content governance: the fact that most harmful material flows through private companies whose business models have historically prioritized growth and engagement over authenticity.
Beyond direct enforcement, the MCMC provides essential technical capabilities to law enforcement agencies investigating deepfake-related crimes. Digital forensic analysis and profiling information extracted from social media platforms enable investigators to trace the origins of synthetic content, identify perpetrators, and build evidentiary records suitable for prosecution. This collaborative approach acknowledges that defeating organized deepfake campaigns requires the combined expertise of communications regulators, cybersecurity specialists, and criminal investigators. For Malaysia and the region, this capability-sharing represents an important step toward institutional capacity building in combating digital manipulation.
The government has also moved to address a related vulnerability: fraudulent advertising masquerading as legitimate commercial content. Platform operators licensed under Malaysian law now face mandatory requirements to verify the identity of all advertisers, drawing on official records from the Companies Commission of Malaysia and other government agencies. This verification regime aims to prevent scammers from creating fake business accounts and purchasing advertising placements that promote investment frauds, romance scams, or product counterfeits. The measure reflects understanding that deepfakes and fraudulent advertising often operate in tandem, with synthetic videos used to establish false credibility for scam operations.
Non-compliance carries substantial financial consequences designed to motivate serious platform investment in enforcement infrastructure. Under the Online Safety Act, social media providers that breach their Risk Mitigation Code obligations face potential prosecution, with convictions carrying fines up to one million ringgit supplemented by additional penalties as high as ten million ringgit. These penalty thresholds represent a significant escalation from previous regulatory regimes and signal that Malaysia intends to treat platform negligence as a serious breach rather than a minor administrative violation. The financial exposure should theoretically incentivize rapid implementation of detection and removal systems across major platforms operating in the Malaysian market.
The policy framework emerging in Malaysia reflects broader Southeast Asian anxiety about artificial intelligence-enabled information warfare. As deepfake technology becomes more accessible and convincing, the region faces mounting risks to democratic processes, financial stability, and intercommunal relations. Countries across ASEAN have begun developing similar regulatory approaches, though coordination remains limited. Malaysia's experience implementing the Online Safety Act will likely inform regional discussions and potentially establish precedents for how developing nations can govern AI-generated content without resorting to crude censorship or completely surrendering oversight to foreign technology companies.
The surge in complaints through mid-2025 suggests that public reporting mechanisms are functioning and that citizens increasingly recognize deepfake content as a distinct category of online harm worthy of regulatory attention. This awareness, while concerning as a barometer of the problem's growth, also creates opportunities for targeted public education campaigns. Authorities can leverage reporting data to identify which communities face heaviest exposure to synthetic content and tailor digital literacy initiatives accordingly. Understanding the vectors through which deepfakes spread—whether primarily through WhatsApp groups, TikTok, Facebook, or messaging apps—enables more surgical interventions than broad-based platform restrictions.
Looking forward, Malaysia's deepfake enforcement efforts will face escalating technical challenges as generative AI tools improve in quality and sophistication. The legislation passed in 2025 provides regulatory tools appropriate for current technology, but the pace of AI development means new threats will emerge faster than policy adjustments can be implemented. The government's decision to build ongoing engagement mechanisms with platform providers and law enforcement agencies rather than relying solely on static legal prohibitions suggests awareness of this challenge. Sustained investment in digital forensic capacity, international cooperation on cross-border deepfake campaigns, and regular review of the Risk Mitigation Code framework will likely determine whether Malaysia can maintain meaningful oversight of synthetic content or faces the scenario that has troubled other nations: where detection lags so far behind generation that regulatory enforcement becomes largely symbolic.
