Indonesia: building a national AI roadmap through cross-government consultation

The problem

Indonesia is the world's fourth most populous country and Southeast Asia's largest economy, with a young, heavily online population. Its government has placed artificial intelligence at the centre of Golden Indonesia 2045, the national ambition to become one of the world's leading economies by the country's centenary. Several structural conditions shape what it can build on.

The government has said it wants to raise research and development spending from around 0.4% of GDP towards 1% and has identified a shortage of skilled workers, with one official putting the gap at around three million digital workers. Responsibility for AI and data sits across several laws and agencies rather than a single AI statute, and personal data is governed by a dedicated law passed in 2022. Digital connectivity varies across the country, with fast mobile coverage reaching most of the population and next-generation networks concentrated in the largest cities. The earlier national AI strategy, set out for 2020 to 2045, also predates the rapid arrival of generative AI from late 2022 and has struggled to keep pace with it.

The solution

To tackle these gaps in a coordinated way, the Ministry of Communication and Digital Affairs, known as Komdigi, led the drafting of Indonesia's first National AI Roadmap, a white paper of around 179 pages setting out the country's AI priorities, infrastructure and financing, with implementation priorities concentrated on the 2025 to 2029 period.

How the document was produced is one of its defining features, and the part most likely to be useful to another government. Rather than being written inside a single ministry, it was assembled by a 443-member task force drawn from government, academia, industry, community groups and the media, spanning 39 ministries and agencies. It was opened for public consultation alongside a companion set of draft AI ethics guidelines, with the consultation closing on 22 August 2025 after more than 400 formal submissions, and it builds on the earlier National AI Strategy for 2020 to 2045, which had been developed by a non-government body of scientists and industry leaders and predated the generative AI surge.

The roadmap is organised around three pillars and three time horizons, short term to 2027, medium term to 2035, and long term to 2045. The first pillar is talent, with targets to produce 100,000 AI professionals a year and to make 20 million citizens AI-literate by 2029. The second is research and innovation, including a shared, cross-sector testing space, or sandbox, where new applications can be trialled under supervision before wider use. The third is infrastructure and data, covering high-performance computing, the specialised chips that AI training depends on, and a national cloud hosted in domestic sovereign data centres so that sensitive data stays under national control. Priority sectors include health, education, food security, public administration, transport and the creative economy, with public services flagged as an early focus, for example, chatbots for citizen queries and data analysis to support policy. The roadmap is aligned with President Prabowo's national development agenda, known as Asta Cita.

Financing is set out as a phased, blended model. A key feature is the role of Danantara, Indonesia's sovereign wealth fund, which the roadmap tasks with designing the financial instruments and a proposed Sovereign AI Fund. The paper recommends a public-private financing model, with early funding directed toward research, public-sector pilots, and computing infrastructure.

The impact

This is a planning document published for consultation, not a deployed system, so the position below describes what it sets in motion rather than measured results.

1. It establishes Indonesia's first coordinated national AI plan.

The roadmap pulls priorities, targets, and financing into a single reference, assembled through coordination across 39 ministries and agencies, replacing a more fragmented picture in which AI and data responsibilities were scattered across overlapping laws and bodies. A legal-sector tracker records it as Indonesia's first comprehensive AI roadmap, released in August 2025 with a companion draft set of AI ethics guidelines, setting a national vision through 2045.

2. It sets concrete, measurable targets.

Rather than a purely aspirational statement, the roadmap commits to specific figures, including 100,000 AI professionals a year and 20 million AI-literate citizens by 2029. These are commitments, not achievements; the document's worth will ultimately turn on delivery against them, and the government has begun standing up training programmes, such as an AI Talent Factory, towards the talent target.

3. It drew broad input, then moved towards becoming binding regulation.

The public consultation closed on 22 August 2025 after more than 400 formal submissions. Rather than remaining a standalone white paper, the roadmap is being formalised as a Presidential Regulation: as of early 2026, the documents were being finalised by the State Secretariat ahead of signing. The government has indicated that the regulation would establish an overarching framework, with individual ministries regulating the details of their own sectors.

Lessons learned

• Cross-sector drafting builds a strategy on input from those who will deliver it. Indonesia produced the roadmap not within one ministry but through more than 443 contributors from government, academia, industry, civil society, and the media, organised into seven specialised groups across 39 agencies, and then opened it to public consultation, which drew more than 400 submissions. The approach gathers the needs and constraints of many sectors before a strategy is fixed.

• Naming constraints alongside target sets ambition against a known baseline. The roadmap pairs its goals with the limits it works against; the report identifies a shortage of skilled professionals, limited research funding, uneven connectivity outside major cities, and risks of misinformation and data breaches, a reading echoed by an AI academic who noted Indonesia remains largely a consumer of foreign AI, citing infrastructure, chip supply and talent gaps. Stating the baseline makes the targets legible: a reader can see what each commitment is measured against.

• A roadmap sets direction, but binding rules are what give it force. Indonesia published the white paper with draft AI ethics guidelines and is converting it into a Presidential Regulation. The sequence shows the distinction in practice: the roadmap directs activity, while the enforceable instrument follows as a separate step, drafted in 2025 and, as of early 2026, being finalised by the State Secretariat ahead of signing.

Launch year: 2025

This case study was written with assistance from artificial intelligence.