Taking down the online degradation

The Star

15-06-2025

WHEN the United States passed the Take It Down Act recently, it was seen as a long overdue move to tackle the rise in 'deepfake porn' or artificial intelligence-generated non-consensual pornography.
The law prohibits anyone from knowingly publishing non-consensual intimate content, whether deepfake or not, and as its namesake suggests, the law also requires social media platforms to take down any non-consensual intimate images within 48 hours upon a report or request from users.
But the problem of deepfake porn is not limited to just the US; in recent years, Malaysia has seen a proliferation of AI-generated explicit content being posted online.
It was previously reported that last year, the Malaysian Communication and Multi-media Commission (MCMC) removed 1,225 postings of explicit content generated by AI as of Dec 1 – a dramatic jump from 186 such cases just two years earlier in 2022.
In April this year, news broke about a case involving artificially-generated obscene images of students in Johor, with Deputy Communications Minister Teo Nie Ching calling for stronger digital safety measures in schools in the aftermath.
Education Minister Fadhlina Sidek issued a similar call this week as she urged for indecent social media pages to be taken down swiftly, following reports of a Facebook group circulating photos of underage students with sexually charged comments.
Her comments came just days after the gazettement of the Online Safety Act (Onsa) 2025 – legislation which is meant to strengthen online protections and hold platforms more accountable.
But while Onsa also introduces a takedown mechanism for harmful content, lawyers and women's rights advocates say its vague language and failure to explicitly cover non-consensual intimate content involving adults leave significant gaps in protection for survivors.
Takedown limitations
Make no mistake, Onsa is a timely update to Malaysia's digital safety laws because it also introduces a takedown mechanism, says Bar Council Cyber and Privacy Laws Committee head Sarah Yong.
The mechanism in Onsa allows users, including victims of online sexual exploitation, to report harmful content directly to licensed social media platforms and the MCMC, she says.
While Onsa considers any child sexual abuse material as 'priority harmful content', Yong says adults depicted in sexual abuse materials such as in deepfake porn may fall under other harmful content categories such as 'obscene' or 'indecent content'.
However, lawyer Daniella Zulkifli points out that Onsa only imposes a duty on licensed service providers to establish a mechanism to make 'prioritised harmful content' inaccessible.
As non-consensual intimate imagery such as deepfake porn involving adults is not included as 'priority harmful content', the takedown provision may not apply to it.
'We need to include non-consensual intimate content as priority harmful content given the nature of it and how damaging it is to someone.
'We do need to improve on our existing laws to allow for a swift takedown of non-consensual intimate content involving children as well as adults,' says Daniella.
Other laws, such as the Sexual Offences Against Children Act, the Penal Code and the Communications and Multimedia Act, can cover non-consensual intimate content, she adds, but none of them compel online platforms to remove such content.
Yong says Onsa is also limited when it involves foreign platforms that are not licensed in Malaysia.
'In this circumstance, the MCMC could instruct a licensed network service provider to restrict the relevant parts of its network service.'
While this would technically cause the content to be inaccessible to all users of the network, Yong says there are ways to circumvent this.
Survivor-centric approach
While Onsa serves its purpose of monitoring and regulating content to a certain extent, Women's Aid Organisation (WAO) advocacy officer Tamyra Selvarajan critiques it for having vague language and failing to address underlying issues such as transparency and accountability.
She also points out that the provisions in Onsa are heavily fixated on content moderation, which she says is not the issue at hand.
'Rather, it is the design of the platform and the lack of user education.'
It is also important to note that deepfake porn is a matter that disproportionately affects women and marginalised communities, says Selvarajan, making this both a digital safety issue and a gender issue.
'There is still a fundamental need to ensure that these authorities are used in a transparent, rights-respecting and gender-inclusive manner,' she says.
Thus, for Malaysia to emulate a law like the Take It Down Act in the US, she says there is a lot to be considered.
'Firstly, there needs to be a survivor-centric approach when it comes to reporting and taking down such content while also safeguarding freedom of expression and privacy.'
This should include survivor- sensitive processes built into the system, Selvarajan says.
'When implementing such policies, intersectionalities such as ethnicity, class and, of course, a woman's standing must be taken into account.'
But Selvarajan warns that though the Take It Down Act is 'somewhat of a win' for Trump's administration, it has been criticised for being vague and overbroad.
Daniella also points out that freedom of speech advocates have flagged the law as lacking critical safeguards against misuse.
'We should take note of the same.'