PETALING JAYA: The Malaysian Communications and Multimeldia Commission (MCMC) has removed more than 400 pieces of deepfake scam content from online platforms between January and mid-August this year Communications Minister Fahmid Fadzil told the Dewan Rakyat.
“From January to August 15, the number of scams involving deepfake misuse of applications or AI technology have amounted to 439 pieces of content requested for takedown, of which 417 have already been removed.
“The majority of these were on Facebook, with 363 requests, followed by TikTok with 59, and Youtube with 17,” he said, responding to Jempol MP, Datuk Shamshulkahar Mohd Deli.
Fahmi also added that MCMC also acted on cases of online harassment and bullying involving AI-generated content.
“For bullying or harassment using deepfake technology, 203 pieces of content were requested for takedown; 104 on Facebook, 98 on TikTok, and one on Platform X.
“These were cases where deepfake technology was used to commit offences against users or viewers of these platforms,” he said.
These actions, Fahmi noted, were part of measures under the amended Online Safety Act 2025 (Act 866), which empowers the government to strengthen protections against harmful online content.
The minister explained that the Act now requires service providers to prepare online safety plans, while new subsidiary instruments being drafted by MCMC will mandate stricter obligations on platforms.
“MCMC is developing regulations and codes that focus on child protection online and risk management.
“It is also studying the need for platforms to implement labelling requirements for content generated by artificial intelligence,” he said, adding that the move aimed to help users identify misleading or manipulative material.
On a separate enquiry by the Jempol MP, Fahmi also responded to a recent viral case involving an elderly couple from Kuala Lumpur who fell victim to a fake AI-generated video.
The couple had been deceived by a manipulated video promoting a supposed new cable car service from Pengkalan Hulu, Perak, to Kedah, leading them to travel nearly 300km and suffer financial losses.
Fahmi said the incident underscored the urgency for platforms to label AI-generated material.
“This is precisely why we believe social media platforms must bear responsibility to automatically label any content generated using AI.
“ We have observed many new applications capable of producing music, videos, and images that appear almost indistinguishable from genuine recordings.
“Once MCMC finalises the subsidiary instruments, we will ensure platforms comply with these requirements. We cannot allow Malaysians to be deceived by AI-generated content or generative AI hallucinations that depict things which do not exist,” he said.
Jerantut MP Khairil Nizam Khirudin also pressed the government on whether heavier punitive measures would be introduced beyond removing links, and whether guidelines would be imposed on media and platforms before publishing AI-generated content.
In his reply, Fahmi said such requirements would fall under Act 866 and must be consistently applied.
“At present, some platforms such as TikTok have already begun allowing content to be labelled as AI generated. However, we do not want this to be done half-heartedly or inconsistently. It must be applied consistently, and it must be mandated under an Act, under the law,” he added.