Simon Butt – Indonesia's old KUHAP was often criticised for making the state and law enforcement too powerful. Coordinating Minister for Law Yusril Ihza Mahendra has called the new Code of Criminal Procedure (Kitab Undang-undang Hukum Acara Pidana, or KUHAP) a new era of law enforcement that is more humane, modern and just.
But nothing could be further from the truth. As the first post in this two-part series argued, the Code has expanded the types of evidence that can be used to pursue citizens and has introduced counterbalances that could prove meaningless.
This post considers how these evidence rules operate in light of the state's coercive powers – at suspect designation, arrest, detention and conviction. It argues that while the Code may, at first glance, appear to make it harder for the state to exercise many of these powers, the new evidence rules, in fact, make it much easier.
From police regulation to statute
The 1981 Code (Law 8 of 1981 on Criminal Procedure) set no explicit numerical evidentiary threshold for the state to name a person a suspect, arrest them or detain them. Over time, a requirement that police have at least two pieces of valid evidence before designating someone a suspect emerged from internal police regulations (Peraturan Kapolri) and, in 2014, from a Constitutional Court decision that formally imposed that threshold.
Law 20 of 2025 on Criminal Procedure is the first statute (undang-undang) to adopt the two-piece standard. Article 90(1) requires a minimum of two pieces of valid evidence before a person can be named a suspect. Article 94 applies the same minimum to arrest, and Article 100(5) to detention (provided other requirements are met). Ministers and senior officials have promoted this as a clear example of the new Code's commitment to rights protection.
The threshold's limits
As the first post explained, Article 235(1) of the new Code also expands the list of admissible evidence from five categories to eight. Physical evidence, electronic evidence and judicial observation (pengamatan hakim) are now standalone categories, and the catch-all in Article 235(1)(h) admits anything lawfully obtained.
For suspects, the problem is that the Code does not require that any of this evidence have probative value at the investigation stage of the criminal process. With that many categories, especially the catch-all one, finding 'two pieces' to support suspect designation, arrest or detention is now trivially easy. A fragment of CCTV paired with a police officer's observation; a text message paired with a witness's second-hand account; a social media post paired with an expert report. Any of these could be more than enough to satisfy Articles 90, 94 and 100, because those provisions test how many pieces of evidence exist, not how strong they are.
This matters because of the leverage the state acquires once those thresholds are cleared. Being named a suspect can ruin a reputation long before trial. Arrest and detention deprive liberty and give police opportunities to extract confessions.
The principal mechanism for challenging the legality of a suspect's designation, arrest, or detention is the pretrial review (praperadilan). Under the 1981 Code, praperadilan was narrow – it could be used to challenge arrest, detention, or the cessation of investigation or prosecution, but not suspect designation. Giving effect to the same 2014 Constitutional Court decision mentioned above, the new Code extends the scope of praperadilan to suspect designation.
Yet while the Code reforms some aspects of praperadilan, it does not change what praperadilan tests. The pretrial judge checks whether the formal requirements have been met – whether the investigator had two pieces of evidence, not whether those pieces are probatively strong. Again, because almost anything can now be used as evidence, it seems likely that suspects will find it difficult to challenge coercive action for evidentiary insufficiency.
The conviction standard quietly shifts
While the new Code applies numerical evidentiary thresholds at the investigation stage, it does the precise opposite at trial.
Article 183 of the 1981 Code required, for a conviction, 'at least two pieces of valid evidence' and the judge's personal conviction that the defendant had committed the offence. Judges often described this as a safeguard that contained both objective (numerical requirements) and subjective (sufficient probative value to convince the judge) elements.
If Article 183 of the old Code provided any protection to defendants by imposing this objective standard, Article 244 of the new Code removes it. The new Code takes away the need for two pieces of evidence. It provides merely that a defendant may be convicted where the court finds the offence 'lawfully and convincingly proven' (terbukti secara sah dan meyakinkan).
Read together with the expanded evidence categories, this inverts the old Code's architecture. The new Code imposes numerical requirements at the stage where the state exercises coercion, and makes that easy to satisfy, but removes numerical requirements at the stage where the state must justify criminal conviction.
A procedural code for this political moment
The new Code should be read with Indonesia's broader trajectory in mind. Scholars and civil society organisations have, for several years, identified a pattern of democratic decline, much of it documented here on Indonesia at Melbourne – the weakening of the anti-corruption commission, the expansion of military roles under the revised TNI Law, and erosions of parliamentary and judicial independence.
Government figures have, for some years, been using the criminal law to target critics, alleging defamation or manufacturing corruption charges. Indonesia now has a procedural code that makes it considerably easier to designate, arrest, detain and convict critics. The new Code now stands as a looming deterrent to anyone with the tenacity to push back.