The Algorithm and the Lunch Tray: Indonesia’s Risky AI Experiment

Indonesia is preparing a presidential regulation that would embed artificial intelligence in key government programmes, including President Prabowo Subianto’s $15 billion free meals plan. The move reflects Jakarta’s effort to catch up with regional AI leaders such as Singapore and Malaysia, but it also raises questions about infrastructure, governance, transparency and whether AI can address deeper implementation problems in public services.

Indonesia is reportedly preparing to integrate artificial intelligence (AI) into some of its most politically important public programmes, including President Prabowo Subianto's $15 billion free meals plan. The proposal raises a harder question. Can AI fix public-service problems that are rooted not only in data gaps, but in governance, capacity, transparency and trust?

According to Reuters, a draft presidential regulation awaiting Prabowo's signature lays out a roadmap for ministries and regional governments to adopt AI from 2026 to 2029. The goal is to support economic growth and accelerate the use of AI in the president's priority programmes. The draft also states AI could lift Indonesia's gross domestic product by 12%, or $366 billion, by 2030, reflecting both Jakarta's aspiration and the risk of overpromising.

Indonesia's Billion-Dollar AI Gamble

Indonesia's move comes as Southeast Asia becomes a more competitive arena for AI and cloud infrastructure. Singapore and Malaysia have moved faster to position themselves as AI development hubs, drawing investment from global technology companies seeking to build the computing backbone needed for cloud and AI services.

Indonesia does not want to remain on the sidelines. According to the reported draft regulation, the government aims to make the country more competitive in AI use regionally and globally. It also reiterates plans for a "sovereign AI fund," to be handled mainly by the country's new wealth fund, Danantara Indonesia, and proposes fiscal incentives for AI researchers as well as efforts to address talent shortages.

However, Indonesia's starting point is uneven. Analysts cited in the draft's context warn that the country is not yet ready to become an AI developer, citing limited infrastructure, a shortage of chips and a lack of AI skills in the workforce. The concern is significant because the difference between adopting AI and developing AI is significant. A government can buy tools, platforms and systems from foreign technology providers. Building the expertise, infrastructure and accountability mechanisms to use them well is a more difficult task.

Can AI Fix a Programme Already Under Fire?

The most politically sensitive part of the plan is the free meals programme, one of Prabowo's flagship initiatives. Under the draft regulation, AI would be used to design region-specific menus, monitor kitchen hygiene, predict food demand, detect irregularities and integrate health data for early warnings of emergencies. Demand forecasting could reduce waste and hygiene monitoring could support food safety. Irregularity detection could help identify procurement or operational problems and health-data integration could improve responses to outbreaks or emergencies.

However, the programme has faced criticism over transparency. Irregularities have been detected in the setting up of kitchens, while safety standards and emergency responses have been criticised after tens of thousands of children suffered food poisoning last year, resulting in the recent dismissal of the programme chief.

In view of this situation, the AI plan seems both attractive and risky. It is attractive because the free meals programme is large, complex and vulnerable to waste. A system that must source food, manage kitchens, serve children and respond to safety incidents across a vast archipelago would benefit from better data and faster oversight.

On the other hand, it is risky because AI works only as well as the institutions around it. If kitchen inspections are weak, procurement records incomplete, health databases fragmented or local officials undertrained, AI may identify some symptoms but fail to address the underlying problems. In the worst case, technology could become a layer of sophistication over a system that still lacks basic transparency.

The draft notes that AI-driven automation has helped organisations improve efficiency and reduce operational costs. However, AI systems also carry costs of their own: infrastructure, software, training, data protection, cybersecurity, monitoring and independent audit.

Big Tech Sees Opportunity, Indonesia Sees a Dilemma

The draft's development also points to the growing role of global technology companies in Indonesia's AI ambitions. Meta Platforms, IBM and Microsoft have reportedly contributed to the draft.

Microsoft has previously announced plans to invest $1.7 billion over several years to expand cloud services and AI infrastructure in Indonesia. Such investment could help the country close infrastructure gaps and help it thrive in the AI era. It could also increase the country's reliance on foreign providers for critical systems.

Indonesia wants to compete in AI, but may initially depend heavily on companies that already dominate cloud, computing and AI services. The proposed sovereign AI fund suggests policymakers understand the need for domestic capacity, but funds and incentives alone will not guarantee a strong local AI ecosystem unless they are matched by skills development, data infrastructure, research capacity and clear procurement rules.

For foreign firms, Indonesia is a large market with significant public-sector demand. For Indonesia, these companies offer technology and investment. The question is how the government balances access to global expertise with the need for local capability, public accountability and long-term control over sensitive systems. This balance is critical when AI is used in welfare and health programmes. The draft also says AI will be used to analyse health checks and support testing for tuberculosis - areas where better data analysis could have public-health benefits, but they also involve sensitive personal information.

The Next Test Is Governance, Not Hype

A serious AI roadmap would need more than ambition. It would need transparent procurement, published technical standards, independent oversight, privacy safeguards, budget clarity and mechanisms for citizens to challenge mistakes.

Indonesia's AI plan should therefore be read as both a development strategy and a governance test. The promise is that AI could make large public programmes more efficient, responsive and accountable whereas the risk is that it becomes a technological answer to problems that are partly institutional.

The free meals programme will be the clearest test. If AI helps improve food safety, reduce waste and detect irregularities, it could strengthen confidence in one of Prabowo's signature policies. If deployed without transparency or capacity, it may deepen doubts about a programme already facing scrutiny.