On May 19, 2026, Amnesty International published "Building Up Imaginary Enemies" (Index ASA 21/0931/2026), a report documenting how state and state-aligned actors in President Prabowo Subianto's first 18 months have weaponized disinformation against critics — branding journalists, academics, and activists as "foreign agents." Amnesty's specific warning lands on a piece of pending legislation: the proposed Law on Countering Disinformation and Foreign Propaganda, which the report says risks deepening an authoritarian trajectory by empowering the state to decide what is "false" or "misleading" without independent adjudication.
That bill does not yet exist in final text. But its architecture is already visible, and it is being built on top of an enforcement machine that is very real.
The strongest case for the bill
It is worth stating the government's argument at its best. Coordinating Minister for Law and Human Rights Yusril Ihza Mahendra confirmed in January 2026 that President Prabowo personally instructed the Law Ministry to draft the legislation, citing "a lot of disinformation and misunderstandings surrounding our national developments." The bill's supporters frame disinformation as an organized, systemic phenomenon — coordinated networks of fake accounts, paid "buzzers," bots, and cross-border influence operations — rather than ordinary citizens speaking their minds. Notably, the government insists the law would target platforms and organized actors, not individuals, and would carry both administrative and criminal penalties.
This is not a frivolous concern. Coordinated inauthentic behaviour is a genuine threat to information ecosystems worldwide, and democracies from the EU to Australia are grappling with how to regulate it. A narrowly drawn statute that targeted demonstrable coordinated manipulation — with transparency obligations and judicial oversight — would be a defensible response.
Where it breaks
The problem is that the bill, as described by its own sponsors, places the determination of truth inside the executive branch. When the government itself decides what counts as "disinformation" or "foreign propaganda," and can impose criminal liability on that basis, the category becomes whatever the state finds inconvenient. Amnesty's report supplies the reason to worry: it counts numerous instances of President Prabowo publicly accusing critics of being foreign-controlled, and documents the human cost. In March 2026, Andrie Yunus, deputy coordinator of the rights group KontraS, suffered an acid attack in Jakarta after months of coordinated online campaigns labelling him a "foreign agent" for leading protests against revisions to the Military Law; four military officers were subsequently arrested.
A law that lets that same state define "foreign propaganda" does not neutralize disinformation — it institutionalizes one side of it. The Jakarta Post reported that rights groups, including YLBHI, warn the bill could be turned against opposition parties, academics, and journalists. The defining feature of legitimate regulation is that the regulator does not get to be the judge of its own critics. This bill, by design, would.
The enforcement layer is already running
The bill would not arrive in a vacuum. It would plug into SAMAN, the content-moderation dashboard the Ministry of Communications and Digital Affairs (Komdigi) rolled out in 2025 under the Electronic Information and Transactions (ITE) Law. SAMAN compresses platform compliance into brutal windows: per government order, platforms must remove flagged "urgent" content within 1×4 hours and standard content within 1×24 hours, or face fines of up to Rp500 million (~US$29,000) per item — or outright blocking. Urgency is classified under Ministerial Decision No. 522 of 2024.
Speed is the point, and speed is the danger. SAFEnet director Nenden Sekar Arum has warned that the fine structure pushes platforms toward over-removal: "The threat of a fine of up to Rp500 million per content could make platforms overly cautious," and the categories SAMAN polices — content "disturbing the public" or "disrupting public order" — are vague enough that "content that is considered disturbing could be legitimate, a vital form of public aspiration." When a platform has four hours to choose between a fine and a takedown, it removes. This is the textbook collateral-censorship problem: liability rules engineered for speed convert private companies into automatic enforcers of the broadest possible reading of the law.
The scale is not hypothetical. Government data compiled by Databoks shows Komdigi processed roughly 2.8 million content takedowns between October 2024 and September 2025 — about 2.1 million of them online gambling. Much of that is uncontroversial. But a pipeline tuned to remove millions of items on short deadlines, with vague categories and steep fines, is precisely the infrastructure a disinformation law would inherit. Add a statute that lets the executive define "falsity," and the takedown machine acquires a new and far more political target set.
A proportionate path exists
Indonesia can fight coordinated manipulation without handing the state a monopoly on truth. The proportionate version of this law would: define disinformation by conduct (coordinated inauthentic behaviour, fake-account networks) rather than by content veracity; require an independent body or court, not a ministry, to adjudicate what is false; build appeal rights and reinstatement timelines into SAMAN; and replace four-hour fine-or-remove ultimatums with windows that allow genuine review. Transparency reporting on every government takedown order should be mandatory.
The open internet has been one of Indonesia's great democratic assets. A government confident in its record does not need to criminalize the contested. It needs to win the argument — in public, before an independent referee, not behind a four-hour takedown clock.