Malaysia's efforts to combat artificial intelligence-generated deception are gaining traction, with authorities successfully removing thousands of deepfake posts from social media platforms in the opening half of 2024. The Malaysian Communications and Multimedia Commission (MCMC) submitted 13,122 removal requests to licensed social media providers between January and June, targeting content that involved image manipulation or deepfake technology. Of these submissions, 12,353 posts—representing 94 per cent of the total—were successfully deleted, according to a parliamentary written reply tabled by the Ministry of Communications.
The crackdown reflects growing government vigilance against digital manipulation technologies that have become increasingly sophisticated and accessible to bad actors. Deepfakes and manipulated media pose distinct threats to Malaysia's information ecosystem, ranging from political misinformation to financial fraud and personal harassment. The high removal success rate indicates that social media platforms are broadly cooperating with Malaysian regulatory authorities, though the sheer volume of flagged content underscores the scale of the problem facing the nation.
Beyond deepfakes, the MCMC has been equally aggressive in pursuing scam-related content. Since January 2022 through the end of June 2024, the commission requested the removal of 275,787 posts involving financial fraud, fake accounts, and impersonation schemes. Licensed service providers successfully took down 262,293 of these posts, translating to a 95 per cent removal rate. This parallel enforcement action demonstrates that manipulation of identity and credentials remains a persistent threat requiring sustained regulatory attention.
A pivotal development came into effect on June 1, when the Risk Mitigation Code took force, establishing new requirements for social media platforms operating in Malaysia. Under this framework, licensed service providers must now clearly label any content that has been generated or substantially altered using artificial intelligence tools. This includes deepfakes, manipulated images, and doctored audio files. The labeling requirement aims to foster digital literacy and help ordinary users distinguish authentic content from artificially modified material, creating friction that may discourage the spread of deceptive content organically.
The Online Safety Act 2025 has introduced additional enforcement mechanisms. During the January-to-June period, the MCMC submitted five removal requests specifically targeting financial scam content under this new legislation, and all five were successfully removed. While the number appears modest compared to historical enforcement under older frameworks, the legislation represents a modernised regulatory arsenal reflecting the evolving nature of online harms that governments must address.
Broad fake news investigations form another critical pillar of Malaysia's anti-disinformation strategy. The MCMC has investigated 574 cases involving false online content under Section 233 of the Communications and Multimedia Act 1998 since January 2022. Of these investigations, 23 have proceeded to court prosecution. Twelve cases have concluded with court judgements, while eleven remain in active trial. The concluded cases resulted in fines totalling RM79,000, and one offender received a six-month custodial sentence after defaulting on payment obligations.
For cases that do not warrant full prosecution, the MCMC has employed a graduated response framework. As of late June, 31 cases had been resolved through compound settlements totalling RM1.22 million, reflecting administrative penalties for less severe violations. An additional 84 warning letters were issued to individuals and entities, serving as a less formal but still meaningful deterrent to further misconduct. Forty-seven cases remained under active investigation, suggesting sustained regulatory momentum, while the remainder were categorised as requiring no further action.
The enforcement response demonstrates Malaysia's multi-layered approach to combating information disorder. Rather than relying solely on prosecution, authorities employ a spectrum of interventions—from content removal and labelling requirements through to administrative compounds, warnings, and ultimately criminal charges. This graduated model recognises that not all false or manipulated content represents equally serious harm, and proportionate responses may be more effective at scale than uniform prosecutorial action.
Regional observers note that Malaysia's regulatory framework increasingly mirrors approaches adopted in other Southeast Asian democracies grappling with similar challenges. The combination of platform cooperation requirements, content labelling mandates, and criminal penalties reflects a global trend toward stricter governance of artificial intelligence-generated deception. However, Malaysia's emphasis on achieving near-universal removal rates from social media companies suggests effective negotiating leverage and genuine compliance by major platforms.
The specific targeting of deepfakes and image manipulation responds to documented real-world harms. Deepfake technology has been weaponised in various jurisdictions to create non-consensual intimate imagery, to impersonate public figures and undermine confidence in institutions, and to perpetrate financially motivated scams. Malaysia's regulatory response acknowledges these risks while attempting to preserve legitimate uses of AI technology in creative industries and other sectors.
Looking ahead, Malaysia's experience suggests that sustained enforcement effort, combined with platform cooperation and clearer labelling standards, can meaningfully reduce the circulation of manipulated content. The 94 per cent removal success rate is notably higher than what many countries achieve, indicating either that platforms are particularly responsive to MCMC requests or that the enforcement requests are precisely targeted. Either way, the data reflects a regulatory environment that is operationally mature relative to the nascent nature of AI governance challenges globally.
However, enforcement statistics alone do not capture the full picture. Removal rates measure only detected content that has been successfully flagged and actioned. The volume of requests suggests that new deepfakes and scams are continuously being generated and posted, creating an ongoing arms race between regulators and those seeking to exploit AI-generated deception for criminal or political gain. Malaysia's regulatory framework must continue evolving to anticipate emerging technologies and tactics.
