JAKARTA, Sept. 5 — Indonesia plans to issue two presidential regulations on artificial intelligence in 2026, marking the first legally binding rules for AI under the Prabowo administration. The move aims to build a sovereign, full-stack AI ecosystem to boost the economy, but a new analysis warns the framework may leave state misuse unchecked. One regulation will address AI ethics, building on guidelines the Ministry of Communication and Digital Affairs (Komdigi) issued in 2023 and 2025.
The second concerns a national AI roadmap, based on a white paper released for public consultation in August 2025. Komdigi Secretary-General Ismail said the regulations will focus on governance and building public trust, ensuring AI development is ethical, transparent and accountable.
Yet a peer-reviewed analysis by Ary Hermawan of the ISEAS – Yusof Ishak Institute, published Sept. 5, cautions that despite promises of safety and accountability, there is little assurance Indonesia will not use AI for state repression. The core problem, Hermawan argues, is that enforcement relies on existing laws with no clear guardrails preventing state actors from misusing the technology. Those laws include article 45A of the 2024 Electronic Information and Transaction Law and section 4 of the 2008 Pornography Act, which could curb AI misinformation and deepfake pornography.
Articles 20 and 22 of the 2022 Personal Data Protection Law require explicit consent for personal data processing, which could block using such data to train AI models — though the law makes no specific mention of AI. The Electronic Information and Transactions Law, however, is notoriously draconian and has been wielded by Indonesia’s political and economic elite to silence critics.
Its selective application has done little to stop a robust influence industry central to elite power struggles, nor has it curbed disinformation in a cyberspace that has grown increasingly authoritarian, marked by cybertroopers and surveillance tools deployed against civil society. Deepfakes were common weapons of information warfare during Indonesia’s 2024 elections, fueling fears AI could become another instrument to repress dissent. Without a strong, independent data protection authority, the Personal Data Protection Law has also proven ineffective.
The government is finalising an independent agency, but questions remain over whether it can investigate state institutions accused of breaching the law. Those concerns are sharpened by state access to biometric data collected through Indonesia’s electronic identification system since 2011. In 2022, TOTM Technologies — the source cuts off mid-sentence — the analysis notes that the lack of independent enforcement leaves state misuse unchecked.
What to watch: whether the incoming regulations name a truly autonomous enforcement body, and how the Prabowo administration responds to concerns that AI governance, as drafted, may protect the state more than the public.
