Malaysia's regulatory authorities have stepped up enforcement against artificial intelligence-generated manipulation and deepfake content, with the Malaysian Communications and Multimedia Commission removing nearly 12,400 problematic posts during the opening half of 2024. Data tabled in Parliament reveals the scale of the effort to police synthetic media across the nation's digital landscape, highlighting both the growing prevalence of manipulated imagery and the effectiveness of coordinated removal campaigns.
The MCMC submitted 13,122 formal requests to licensed social media platform providers requesting the deletion of content created through image manipulation or deepfake technology between January 1 and June 30 this year. Of these requests, 12,353 posts—representing a 94 per cent compliance rate—were successfully removed by the platforms. The high removal rate demonstrates the cooperation between Malaysian regulators and international social media companies in addressing synthetic media threats, though the sheer volume of problematic content suggests deepfakes remain a persistent challenge in the Malaysian digital ecosystem.
The deepfake crackdown forms part of a broader regulatory strategy targeting multiple categories of harmful online content. During the identical six-month window, authorities requested removal of 275,787 posts related to scams, including fraudulent accounts and impersonation schemes. A complementary 95 per cent of these scam-related posts were successfully taken down, indicating that platform providers are broadly aligned with Malaysian regulatory objectives and maintain infrastructure capable of executing mass content moderation at scale.
Understanding the Malaysian enforcement landscape requires awareness of the recent regulatory overhaul governing artificial intelligence on digital platforms. The Risk Mitigation Code, which commenced on June 1 this year, now mandates that licensed platform providers apply labels to content that has been generated or substantially altered using AI tools. This requirement encompasses deepfakes, manipulated images, and synthetic audio. The labelling regime represents a shift toward transparency rather than outright prohibition, enabling users to make informed judgments about content provenance while allowing researchers and authorities to track the proliferation of AI-generated material across social networks.
The government has also deployed newer legislative instruments to address online harms. Under the Online Safety Act 2025, authorities submitted five separate removal requests specifically targeting financial scam content during the January-to-June period. All five requests resulted in successful content deletion, suggesting that the dedicated scam-fighting provisions are proving effective, though the relatively modest number of requests indicates either limited platform deployment of this mechanism or the adequacy of existing tools for most cases.
Enforcement activity extends beyond mere content removal into criminal prosecution. Between January 2022 and June 2024, the MCMC investigated 574 cases involving false or misleading online content under Section 233 of the Communications and Multimedia Act 1998. Of these investigations, 23 cases proceeded to judicial prosecution. Twelve cases have concluded, resulting in courts imposing combined fines totalling RM79,000. Notably, one offender received a six-month custodial sentence after declining to settle an imposed fine, indicating the courts' willingness to apply custodial penalties in serious instances where fines alone prove insufficient deterrent.
The prosecution pipeline remains substantial, with eleven cases still undergoing trial as of the parliamentary response date. An additional 31 cases have received formal compounds—predetermined settlement amounts—totalling RM1.22 million collectively. The MCMC issued 84 warning letters in cases deemed unsuitable for prosecution but warranting formal notice. Forty-seven cases remain under active investigation, while the remainder were classified as requiring no further action, either because they fell outside regulatory scope or lacked sufficient evidence of breach.
The regulatory scrutiny has extended to prominent news-adjacent Facebook accounts. When queried about the HarakahDaily Facebook page, the ministry confirmed that no First Information Report had been registered as of June 30, meaning no formal criminal complaint had been lodged against the account holders. However, authorities signalled readiness to pursue enforcement if future investigations revealed breaches of applicable law or the platform's community standards. This cautious stance reflects the tension between combating misinformation and protecting editorial freedom in Malaysia's contested media landscape.
For Malaysian businesses and citizens, these enforcement statistics carry practical implications. The aggressive removal of scam-related content and deepfakes signals that regulatory resources are directed toward protecting online consumers and users from identity theft, financial fraud, and reputational manipulation. Simultaneously, the labelling requirements for AI-generated content will reshape how Malaysian social media users encounter algorithmically-amplified material, potentially enhancing their ability to distinguish between authentic and synthetic content.
The regional context further sharpens the significance of Malaysia's approach. As Southeast Asian economies increasingly grapple with cross-border digital threats and election-season misinformation campaigns, Malaysia's enforcement achievements demonstrate that coordinated action between government regulators and platform operators can yield measurable results. The 94 per cent removal success rate compares favourably with international benchmarks and suggests that Malaysian platforms maintain sophisticated content moderation infrastructure.
Looking forward, the transition from ad-hoc takedown requests toward systematic AI labelling represents a maturation of Malaysia's digital governance. Rather than purely reactive enforcement, the Risk Mitigation Code shifts responsibility toward platform providers to implement preventative measures. This regulatory evolution aligns with international discussions about algorithmic transparency and synthetic media governance, positioning Malaysia among nations actively shaping how artificial intelligence-generated content is managed in the digital public sphere.
The parliamentary data also hints at persistent challenges. The ongoing investigation of 47 cases and the eleven trials still underway indicate that many deepfake and scam incidents remain unresolved. Furthermore, the prosecution rate—23 cases out of 574 investigations—underscores the resource constraints facing regulators tasked with policing an exponentially growing volume of online content. As deepfake technology becomes more accessible and convincing, maintaining regulatory effectiveness will demand continued investment in investigative capability and prosecutorial expertise.
