MY govt in talks with TikTok to restore accounts of 18 AI-blocked media outlets, says Fahmi

The Malaysian government is reportedly in talks with TikTok to restore 18 accounts belonging to media outlets. The accounts were removed as a result of artificial intelligence (AI) moderation.

Speaking to media, communications minister Fahmi Fadzil said that the accounts were banned due to the outlets' coverage of an alleged sexual assault of a young girl at a mosque in Selangor last Friday (21 Feb), reported The Star.

The minister said that AI can miss the mark from time to time and that it does not understand that reports made by media outlets are different from content produced by social media users. 

Don't miss: TikTok removes 1.5 million Malaysian videos, enhances safety guidelines

In addition, Fahmi had asked TikTok for a discussion to give leeway or establish a different status to TikTok accounts belonging to media outlets so that such incidents will not occur again. 

According to TikTok's community guidelines, child sexual abuse material is not permitted. This includes screenshots or clips from the original material. The same guidelines apply even if nudity or sexual activity is not depicted. 

This isn't the first time TikTok has removed content from Malaysia that has violated its safety guidelines. In January last year, the video-sharing app removed 1.5 million videos from Malaysia that violated its content moderation and community protection laws. 

In 2023, TikTok removed a total of 1,555,199 videos from Malaysia. These videos went against TikTok's policies such as integrity and authenticity, mental and behavioural health, privacy and security, regulated goods and commercial activities, safety and civility, sensitive and mature themes and youth safety and well-being. 

Of the 1,555,199 videos removed, 96.8% of it were proactively removed before it could be reported by viewers while 77.6% of the videos were removed before it could garner any views. Additionally, 90.8% of the videos were removed within 24 hours of posting. 

Meanwhile, the MCMC has been busy with its own content moderation, with the ministry removing a total of 1,225 items of AI-generated explicit contentas of 1 December this year, compared to 186 in 2022. 

In light of the increase in alarming content, MCMC proposed an amendment to the Communications and Multimedia Act 1998 (Act 588), where anyone found to be distributing explicit content for commercial purposes could face up to five years in jail, a fine of up to RM1 million.

Related articles:  
More than 60,000 fraudulent content removed from social media in 2024 
MY unlikely to be affected by Meta's end to fact-checking, says Fahmi   
MY will not be implementing social media limit law, says Fahmi