MCMC has filed over 345,000 takedown requests for harmful online content this year, with platforms urged to improve AI-based moderation.
PETALING JAYA: The Malaysian Communications and Multimedia Commission (MCMC) has spent at least 19.7 years of manpower filing takedown reports for harmful content since Jan 1, Communications Minister Datuk Fahmi Fadzil told the Dewan Rakyat today.
He said MCMC had identified and requested the removal of 345,712 items of harmful online content between Jan 1 and July 1, with each report taking an average of 30 to 45 minutes to file with the relevant platform.
Based on the lower estimate of 30 minutes per report, Fahmi said the workload amounted to the equivalent of 19.7 years in man-hours.
“If we take the figure of 345,712 and multiply it by 30 minutes for each application, it means that from Jan 1 until yesterday, MCMC has used the equivalent of 19.7 years in man-hours to file reports on content believed to be harmful,” he said.
Fahmi said 91% of the harmful content identified involved gambling or scams, which he described as repetitive content produced by criminals and promoted through the same platforms.
He said the larger issue was that harmful content should be dealt with by the platforms themselves, rather than relying heavily on manual reporting by MCMC or complaints from the public.
He said the Online Safety Act 2025 and its subsidiary legislation were intended to ensure the country’s digital ecosystem, including social media platforms, became safer through action taken automatically by the platforms themselves.
Fahmi said MCMC was studying several measures to improve its capacity, including the use of agentic artificial intelligence (AI) to manage online harm more efficiently.
He said platforms also needed to improve their own use of AI to detect and remove harmful content.
Citing reports involving ByteDance, which operates TikTok in Malaysia, Fahmi said some employees who were recently laid off were believed to have been from the trust and safety division.
He said ByteDance had previously used AI to examine breaches of its community guidelines, but platforms were still not sufficiently capable of using AI to remove harmful content.
“So the platforms need to use AI, and MCMC will also use AI to manage issues involving online harm,” he said.
Fahmi was replying to Rodziah Ismail (PH-Ampang), who asked whether MCMC’s capacity, manpower and technical expertise would be strengthened as regulatory instruments under the Online Safety Act come into force.
Earlier, Fahmi said MCMC had developed 10 subsidiary instruments under the Act, covering regulations, codes and guidelines.
He said four subsidiary regulations came into force on Jan 1, followed by two more on July 1, while the final subsidiary instrument, relating to certain characteristics of private messaging features, is undergoing public consultation from June 19 to July 20 before being finalised.
