Content takedown disputes in Indonesia and a civil society coalition weighing legal action

In October, the Indonesian digital rights group SAFEnet reviewed two years of digital controls under the current government and recorded at least 73 pieces of content taken down between April and September 2026, most of them critical of the government. SAFEnet and other groups are also gathering potential plaintiffs to assess legal action against the Ministry of Communication and Digital Affairs. The ministry denies asking platforms to restrict content about forest fires, and in September the Constitutional Court ruled that the government must give notice with reasons before cutting off access. Readers outside Indonesia are not directly affected.

On 21 October, the Indonesian digital rights group SAFEnet published a report on two years of digital controls under the current president and vice president, arguing that digital space is increasingly becoming a site of repression, surveillance and censorship. By SAFEnet's count, 5,549 digital rights violations were recorded from 2025 to mid-2026, a figure we could not check against any independent source. On 23 October, a legal team formed by SAFEnet with a journalists' association, legal aid groups and others opened a desk to hear from people whose content has been restricted. The restrictions apply inside Indonesia, and readers elsewhere are not directly affected.

According to SAFEnet's report, it and the media monitoring group Remotivi recorded at least 73 pieces of content taken down between April and September 2026, most of them criticising the president, the free nutritious meals programme or disaster response. A separate SAFEnet statement of 5 October gives the period as April to August, and says 90.4% were moderated or removed at the request of the Ministry of Communication and Digital Affairs (Komdigi). According to the same statement, the notices platforms sent on the ministry's request never stated which law had been broken. We found no public response from Komdigi to the figures of 73 items and 90.4%.

The National Commission on Human Rights (Komnas HAM) questioned Komdigi after receiving complaints of "internet restrictions" and "social media restrictions" during protests. On 13 October a commissioner relayed Komdigi's account that it removed 486 pieces of suspected illegal content, based on its online patrols, on 27 August, the day of large-scale protests. Komdigi also denied that it restricted the internet. The commission is still examining the legal basis for those removals.

Komdigi's own figures show about 4.2 million pieces of illegal content removed between October 2024 and April 2026, mostly online gambling and pornography. These are the ministry's own numbers, which we also could not check independently. The 486, the 4.2 million and SAFEnet's 73 are three different counts and cannot be added together or compared.

In mid-October, public figures said their posts about forest fires were reaching fewer people. Komdigi's deputy minister and then the minister denied that the ministry had asked platforms to remove or restrict information about the fires. In a video statement on 21 October, the minister also said the ministry is pushing for accountability and transparency in platform algorithms, and asked users to appeal to the platform first and then report to the ministry.

The legal team has not said in its press release what it would sue over or in which court. According to the release, filling in the form does not make someone a plaintiff, and that the choice of plaintiffs depends on evidence, harm and standing (whether a person is entitled to bring the case). The team also asks people to include screenshots of moderation notices from the platform, links to the content, and their correspondence with the platform.

Perspective

Ministerial Decree 127/2026, signed by Komdigi in March, classifies disinformation and hate speech as content that causes public unrest, and requires platforms to carry out an access cut-off order from the minister within four hours of receiving it. We could not find the decree itself, so its content here follows Kontan's quotations. According to the director for digital space control interviewed by Kontan, reported content is first reviewed jointly with the relevant ministries and agencies. Objections to the government are filed by the platform through SAMAN (the system the ministry uses to send orders to platforms and track compliance), and the ministry separately encourages platforms to offer users a way to object.

The ministry's reasons are set out in the considerations of the decree: disinformation and hate speech can cause public unrest, panic and conflict between groups. Before the decree was issued, the director general of digital space supervision also said in September 2025, when discussing SAMAN, that takedowns are not meant to silence criticism. In April, the Alliance of Independent Journalists (AJI) called for the decree to be revoked, arguing that terms like "causing public unrest" are too vague.

According to a commentary published by the University of Melbourne's Indonesia research outlet, a March Instagram post by the independent outlet Magdalene, which introduced an investigative report on an acid attack, was restricted inside Indonesia under this decree and the restriction was lifted a few days later. The commentary quotes the director general as saying the post's narrative and title could mislead the public and trigger distrust of state institutions, and that Magdalene is not on the Press Council's list of verified outlets. A Press Council member said Magdalene is a legitimate news organisation and that Press Council verification is not a legal requirement.

In its ruling of 29 September (311/PUU-XXIV/2026), the Constitutional Court declared the provision of the Electronic Information and Transactions Law that authorises the government to cut off access conditionally unconstitutional: the government must first issue a digital or written notice with reasons before it cuts off access or orders a platform to do so. The court did not accept the petitioners' argument that prior court approval should be required. As of 29 October, we had not obtained any public statement from Komdigi on the ruling, and found nothing on whether it has changed how it gives notice. SAFEnet calls the ruling a first step towards reform, and argues that because the ministry is both regulator and enforcer, notice alone does not guarantee that restrictions are proportionate.

OONI (Open Observatory of Network Interference, censorship measurements run by volunteers in many countries) and Cloudflare Radar (traffic monitoring) measure whether websites can be reached and overall traffic, so the removal of a single post or a drop in reach does not show up in their data. OONI's website connectivity measurements show no spike in anomalies on 27 August, and Radar has no outage annotation for Indonesia from late August to early September (checked on 29 October). OONI's coverage depends on where volunteers are and is uneven across telecom operators, so regional or operator-level interference cannot be ruled out.

For people whose content has been restricted or removed by a platform in Indonesia, the minister's suggested route is to appeal to the platform first and then report to the ministry. Keeping screenshots of notices and links to the content gives a basis for any later appeal or request for legal help (our suggestion, drawing on the types of evidence the legal team lists).

We are a community based in Taiwan. Elsewhere in Asia, who can order content removed and whether a court is involved are settled differently. In Thailand, the urgent procedure in a notification published in July has officials petition a court, with evidence, to suspend dissemination or remove online data.

In Taiwan, Article 46 of the Protection of Children and Youths Welfare and Rights Act requires platforms, once the competent authority notifies them that content is harmful to children, to restrict children's access to it or take it down. The notice goes to the platform, the article sets out no court procedure, and its scope is limited to child protection.

In Hong Kong, Schedule 4 of the Implementation Rules for Article 43 of the Law of the People's Republic of China on Safeguarding National Security in the Hong Kong Special Administrative Region (2020) lets the Commissioner of Police, with the approval of the Secretary for Security, authorise a designated officer to act on electronic messages likely to constitute an offence endangering national security. The officer may require the person who published a message to remove it by a deadline, or require the platform to take a disabling action (removing the message or restricting access to it). If the platform fails to comply or that is not practicable, the requirement can go to a hosting service provider, and failing that, a network service provider. The rules do not require a court first.

Under the same rules, the publisher may receive a removal requirement directly and in writing, but Schedule 4 contains no provision on giving the publisher reasons or a route of appeal. Indonesia's ministerial decree and access cut-off provision let an executive body give orders directly, as in Taiwan and Hong Kong, while Thailand has officials go to a court. The Constitutional Court has now added prior notice with reasons.

When comparing approaches, it helps to ask who can order content removed, whether a court is involved, and whether the person affected receives a notice with reasons and can object. The intended effect of the policy belongs in the comparison too: whether there are published figures to test how well it is curbing disinformation, gambling and pornography.

Get new stories by RSS, newsletter or Bluesky