Malaysia's efforts to stem the spread of digitally manipulated content have resulted in the removal of 12,353 deepfake-related posts during the first half of 2024, according to parliamentary replies tabled by the Ministry of Communications. The Malaysian Communications and Multimedia Commission (MCMC) filed 13,122 requests for content removal across social media platforms between January and June, with licensed service providers successfully taking down 94 per cent of the flagged material.
The scale of enforcement action underscores the growing sophistication and frequency of deepfake attacks in Malaysia's digital ecosystem. Deepfake technology, which uses artificial intelligence to create convincing but fabricated audio or video content, has emerged as a potent tool for malicious actors seeking to deceive the public, damage reputations, or influence public opinion. The high success rate in takedowns suggests that platform providers are responding swiftly to government requests, though the volume of removals also indicates the persistence of the problem.
The regulatory framework governing platform accountability tightened considerably with the introduction of the Risk Mitigation Code on June 1, 2024. Under this new requirement, licensed service providers must now label any content that has been generated or substantially altered through artificial intelligence tools. This labelling mandate extends to deepfakes, digitally manipulated images, and synthetic audio, creating a transparency layer designed to help users identify suspect content before engaging with it. The policy represents a shift towards placing responsibility on platforms themselves to identify and flag AI-generated material proactively.
Beyond deepfakes, the MCMC's enforcement activities have expanded to address the related threat of scam-related content. Between January 2022 and June 2024, the regulator requested removal of 275,787 posts involving scams, fake accounts, and impersonation schemes. Of these requests, 262,293 posts were successfully taken down, achieving a 95 per cent removal rate. The comparable success rates for both deepfakes and scam content suggest that the MCMC's requests are meeting minimal resistance from platforms, though critics argue that such reactive measures address symptoms rather than underlying problems.
The Online Safety Act 2025 has introduced additional enforcement mechanisms specifically targeting financial scams. Between January and June 2024, five requests for content removal under this legislation were submitted, and all targeted content was successfully removed. Though the number appears modest, the 100 per cent success rate indicates that platforms are fully compliant with removal orders under the new act, possibly because of heightened regulatory scrutiny or the severity of financial fraud cases.
Legislative action against false online content has proceeded through the courts, though at a measured pace. The MCMC investigated 574 cases involving false online information under Section 233 of the Communications and Multimedia Act 1998 during the two-and-a-half-year period from January 2022 to June 2024. Of these, 23 cases reached prosecution, with twelve concluded and eleven still in trial. The courts imposed RM79,000 in total fines across the concluded cases, while one offender served a six-month prison sentence for non-payment of a fine.
The enforcement data reveals the breadth of the MCMC's response infrastructure. Beyond prosecutions, the regulator has issued 84 warning letters and offered 31 compound settlements totalling RM1.22 million. Forty-seven cases remained under investigation as of June 30, while others were classified as requiring no further action. This tiered enforcement approach—combining prosecution, compounds, warnings, and investigation—reflects an attempt to calibrate responses according to the severity of violations and the likelihood of repeated offences.
The question of high-profile cases has also surfaced in parliamentary discourse. The MCMC clarified that no First Information Report had been filed regarding the HarakahDaily Facebook account as of the end of June 2024, though the regulator warned that appropriate action would be taken if any posts breached existing laws or platform guidelines. This measured stance on politically sensitive accounts suggests caution in applying enforcement mechanisms that might invite accusations of selective regulation, a persistent concern in Malaysia's media landscape.
For Malaysian readers and Southeast Asian observers, these developments carry significant implications. The region's vulnerability to coordinated disinformation campaigns has been well documented, and Malaysia's position as a multicultural, multilingual nation makes it particularly susceptible to deepfakes targeting religious, ethnic, or political divisions. The MCMC's proactive enforcement and the new regulatory framework represent attempts to raise barriers to manipulation, yet questions remain about whether takedown rates and fines deter sufficiently sophisticated actors with political or financial motivation.
The labelling requirement introduced under the Risk Mitigation Code also reflects a global shift toward transparency-based regulation. Platforms are increasingly expected to disclose the origin and nature of content rather than simply removing it. This approach assumes that informed users will make better judgments about manipulated media, though research on digital literacy and the effectiveness of labels in reducing engagement with false content remains inconclusive.
The role of artificial intelligence in creating deepfakes continues to evolve faster than regulatory frameworks can adapt. While Malaysia's current measures address detection and removal, questions persist about prevention. The cost of creating convincing deepfakes has fallen dramatically, and the technology's dual-use nature—legitimate applications in entertainment and education alongside malicious ones—complicates regulatory design. Stakeholders from civil society, academia, and the technology sector have called for greater investment in media literacy and AI ethics education to complement enforcement efforts.
As Malaysia approaches the 2025 elections and beyond, the credibility of visual and audio evidence in public discourse will likely become even more contested. The MCMC's enforcement actions and the regulatory infrastructure now in place provide a foundation for response, yet the challenge of maintaining public trust in authentic information while combating malicious manipulation remains ongoing. Policymakers and platform operators will need to balance transparency, speed of enforcement, and due process protections as deepfake technology becomes more prevalent and harder to distinguish from genuine content.
