The Contradiction at the Heart of Jakarta's Disinformation Regime
Indonesia's Ministry of Communication and Digital Affairs — Komdigi — has a strong answer to online disinformation: mandatory takedowns within four hours. That is the deadline built into Ministerial Decree No. 127/2026 (KM 127/2026), issued March 13, 2026, which governs the SAMAN content moderation compliance system. Platforms that miss the window face financial penalties of up to 500 million Indonesian rupiah per piece of non-compliant content, and repeated failures can trigger a full service block.
The underlying concern is legitimate. Indonesia has documented vulnerabilities — significant portions of its internet user base encounter politically motivated false content daily, low digital literacy is widespread outside major urban centres, and the country has a documented history of ethnic and religious incitement cascading through social media. A regulation that forces platforms to actually act — rather than defer indefinitely — sounds like exactly what a responsible digital regulator should want.
The problem is what KM 127/2026 has been used for — and what it has conspicuously failed to do.
What Amnesty Documented
On May 19, 2026, Amnesty International published Building Up Imaginary Enemies: Misinformation, Disinformation and 'Foreign Agent' Allegations in President Prabowo's Indonesia (Index ASA 21/0931/2026). The report documents coordinated campaigns by hundreds of accounts "acting in sync" across Instagram, Facebook, X, TikTok, and YouTube — many presenting themselves as affiliated with the Indonesian military — systematically branding activists and journalists as foreign agents (antek-antek asing).
The campaigns are not spontaneous. They deploy identical videos, graphics, and messaging across platforms in rapid synchronized bursts. Content that the platforms' own community standards would classify as coordinated inauthentic behavior has stayed online for months — in some cases more than a year — without removal under KM 127/2026.
The Ministry's four-hour clock, it turns out, ticks only when Komdigi issues a takedown request. There is no symmetric obligation on the SAMAN system to act against coordinated state-aligned influence operations.
President Prabowo Subianto has publicly accused critics of being "foreign lackeys" or foreign agents at least 25 times since taking office in October 2024, according to the Amnesty report. That framing translates directly into online targeting. Activists who lead protests are cast in military-affiliated content as foreign-funded saboteurs; those campaigns precede or accompany physical threats and, as the next case shows, physical violence.
The Andrie Yunus Case
The most thoroughly documented case involves Andrie Yunus, a coordinator at KontraS — the Commission for the Disappeared and Victims of Violence — who helped lead protests against a 2025 revision to the Military Law permitting active-duty officers to hold positions in 14 civilian government institutions.
KontraS offices were placed under surveillance by army vehicles beginning in early 2025. Online, accounts linked to military units posted videos titled "Indonesia is in danger, foreign lackeys are attacking" featuring Yunus by name. On March 12, 2026, at approximately 11:37 PM in Central Jakarta, two men on a motorcycle threw acid at him as he rode home from recording a podcast on military reform. Chemical burns covered approximately 20% of his body; he was blinded in one eye.
Four officers from the TNI's Strategic Intelligence Agency (BAIS) — Navy Sgt. Edi Sudarko, First Lt. Budi Hariyanto Widhi Cahyono, Capt. Nandala Dwi Prasetia, and Air Force Lt. Sami Lakka — were arrested and convicted. A Jakarta Military Court sentenced them on June 10, 2026 to between 18 months and three years. Human rights organizations including FORUM-ASIA and OMCT condemned the sentences as inadequate for an act of premeditated political violence.
What makes the KM 127/2026 angle particularly striking: a TikTok post targeting Yunus as a foreign agent was published April 5, 2026 — three weeks after the acid attack, and during active criminal proceedings against the military perpetrators. It was not removed within four hours. Meanwhile, when the digital outlet Magdalene published an Instagram post in early April documenting the attack with CCTV footage of suspects, Meta geo-blocked it following a Komdigi request under KM 127/2026. Komdigi's stated rationale: the content "could potentially mislead the public" and raise "groundless suspicions about state institutions." The restriction was lifted only after sustained press freedom advocacy. Other outlets covering the same story — Tempo, Kompas, Tirto — faced no comparable action.
The disinformation stayed up. The journalism about it came down.
A Bill That Would Expand the Same Logic
The government is not pausing to audit this outcome. In January 2026, Coordinating Minister Yusril Ihza Mahendra announced the RUU Penanggulangan Disinformasi — the Countering Disinformation and Foreign Propaganda Bill — on direct instruction from President Prabowo. The bill is not yet in Indonesia's formal legislative priority lists for 2026 or the 2024–2029 parliamentary term, but it could be fast-tracked under political pressure.
The draft framework would apply to digital and non-digital platforms, social media, broadcasters, and "organized actors" including paid political influencers. The government frames it as filling gaps in the 2024 Electronic Information and Transaction Law (UU ITE) and the 2002 Broadcast Law. Critics at Hivos and other civil society organizations note that its definitions of "disinformation" and "foreign propaganda" are broad enough to reach legitimate journalism, diaspora advocacy, and foreign-funded civil society — precisely the categories already being targeted by coordinated military-linked online campaigns.
What Proportionate Design Requires
None of this means Indonesia lacks a genuine disinformation problem. It does. But a takedown mandate with no independent oversight of the requesting authority, and no symmetric enforcement obligation against coordinated state-aligned influence operations, is not anti-disinformation law. It is a censorship mechanism with better branding.
Proportionate reform would require at minimum: an independent review body empowered to audit both government takedown requests and platform compliance with coordinated inauthentic behavior policies; published Komdigi transparency reports on every content removal request by category; and a statutory definition of "disinformation" that explicitly excludes verified reporting on public officials and documented evidence of state conduct.
The Amnesty report found that only TikTok responded to its inquiries about the documented influence campaigns; Meta, X, and YouTube did not. Platform accountability matters — but when platforms selectively act against civil society while leaving state-affiliated campaigns in place, the failure is not only one of platform governance. It is a government that has designed its laws to produce exactly that asymmetry.