What Komdigi said
On 23 October 2026, Deputy Communications and Digital Minister Nezar Patria told reporters in Bogor that his ministry is closely watching AI-generated mini-dramas on social media and over-the-top (OTT) apps. He said they have been infiltrated by online gambling, pornography and sexual violence. He described social media as only a "gateway" to other platforms that sell paid content with almost no censorship. Komdigi is working with platforms on a shared dashboard to monitor content that may break the law. It also welcomes input from the Film Censorship Board (LSF), and it is examining services that are not registered in Indonesia but use social media to reach users (Investortrust).
The strongest case for acting is plain. Short vertical dramas are cheap to produce with generative AI, and sexual violence and gambling promotion cause real harm. Indonesia also has a legitimate child-safety interest. A ministry that did nothing while such material reached minors would deserve criticism.
The tools already exist
The reported plan is mostly monitoring and coordination, not a new rule. That is the right starting point, because Indonesia already has a dense stack of platform obligations.
Ministerial Regulation 5 of 2020 on private-scope electronic system operators (PSE), promulgated on 24 November 2020, requires registration. It also covers moderation of electronic information and takedown duties for prohibited content, and it requires operators to give access to systems and data for supervision and criminal law enforcement (Komdigi JDIH). Those obligations are what let Komdigi go after the services Patria described as unregistered.
Child protection is covered separately. Government Regulation 17 of 2025 (PP Tunas) on electronic system governance for child protection is listed in the national regulations database (BPK). Its implementing rule, Ministerial Regulation 9 of 2026, requires platforms to restrict accounts for users under 16. Komdigi said in April 2026 that seven of the eight platforms it classed as high-risk had submitted compliance commitments (RRI).
So gambling and pornography in AI mini-dramas are not a regulatory gap. They are an enforcement target.
Where the risk lies
The PSE regime has a documented record of blunt use. SAFEnet called Regulation 5 a repressive instrument that could conflict with human rights. It argued that any access termination should meet a three-part test of legal basis, legitimate purpose and necessity (CNN Indonesia). In 2022 the government blocked unregistered services including PayPal, Yahoo and Steam. Many ordinary users were affected, and several services were later unblocked after public pressure (Human Rights Monitor).
That history shapes how we should read the new initiative. A shared dashboard is only as good as its criteria. If it flags content that is illegal under existing statutes, such as gambling promotion, pornography and sexual violence, it is a useful triage tool. If the criteria drift toward vague categories of harmful content, a platform that wants to avoid blocking will over-remove, and legitimate creators will absorb the cost.
The LSF question
The more consequential choice is how far film-style censorship should reach. Patria also said AI lets anyone produce violent and pornographic scenes that look like real footage, and that the LSF needs stronger competence while keeping human judgment central. That may be reasonable for paid OTT services that behave like cinema. Pre-classification does not scale to user-generated social video, where huge volumes of clips are uploaded daily. Applying it there would slow every Indonesian creator who experiments with AI video in order to catch a small share of bad actors.
The gateway structure Patria describes suggests a narrower approach:
- Enforce PSE registration against the paid destination apps. They are the ones with near-zero filtering, and they are where payment happens.
- Use the dashboard for content that is already illegal, with a published definition of what is flagged and a route for creators to contest wrongful flags.
- Target the money and the funnel. Gambling links and payment channels are easier to identify than a mood or genre.
- Reserve LSF-style classification for services that choose to operate as film distributors, not for every upload.
What to watch
Two things will show which way this goes. The first is whether Komdigi publishes the dashboard's criteria and the number of takedowns it triggers. The second is whether blocking is used against specific unregistered apps or against whole categories of service. Regulation 5 and PP Tunas give Komdigi enough authority to deal with the problem as described. The open question is whether it will use that authority with precision.
Indonesia can protect minors and still be a place where AI video creators want to build. That outcome depends on targeted enforcement of laws that already exist, with transparent criteria, not on a broader censorship layer over every platform.