KUALA LUMPUR, July 2 — Communications Minister Datuk Fahmi Fadzil said the government is exploring the use of agentic artificial intelligence (AI) as the Malaysian Communications and Multimedia Commission (MCMC) grapples with a growing workload in online safety enforcement.

He said the scale of enforcement has surged significantly, with MCMC identifying and requesting the removal of 345,712 pieces of harmful online content since January 1.

“Each piece of content requires supporting documentation and justification explaining why it is deemed harmful. According to MCMC, on average each complaint takes between 30 to 45 minutes to process and file with the relevant platforms.

“If we take the figure of 345,712 and multiply it by an average of 30 minutes per case, MCMC has utilised a total of 19.7 years of man-hours of working time to process and file content deemed to be harmful. This demonstrates the significant workload involved,” he told the Dewan Rakyat today.

He was responding to a supplementary question from Ampang MP Rodziah Ismail, who asked whether the government would increase MCMC's manpower, technical expertise and funding to match its expanded responsibilities following the new online safety laws.

Fahmi also said he had instructed MCMC to review its processes to ensure human resources are not overstretched or wasted.

He said the issue did not lie solely with the regulator, but also with digital platforms.

“For example, of the total figure I mentioned, 91 per cent relates to gambling and scams,” he said, adding that such content is often repeatedly generated and promoted through paid advertising systems on platforms.

Fahmi said enforcement efforts have been hampered by limited action from platforms despite repeated complaints.

He added that under the Online Safety Act and related subsidiary regulations, the government aims to strengthen platform accountability to ensure harmful content is removed more efficiently and automatically.

On the use of AI, Fahmi said both regulators and platforms must adopt the technology to manage online harms more effectively.

“Regarding the second question on AI usage, I saw a post on Threads yesterday noting that ByteDance, which operates TikTok in Malaysia, has recently reduced staff in its Trust and Safety division.

Previously, ByteDance itself used AI to detect violations of community guidelines. However, we observe that platforms still struggle to effectively use AI to remove harmful content,” he added.

Therefore, Fahmi said both sides must use AI, with platforms deploying AI systems while MCMC also adopts AI to manage harmful online content.