AI and the New Infrastructure of Development

For now, Indonesia is positioned merely as an AI user. AI tools have penetrated workplaces, academic institutions, media platforms, commercial entities, and governmental bodies. However, adopting external technological solutions remains fundamental distinct from establishing domestic capabilities to construct, guide, and regulate them. Currently, Indonesia's technical, political, and intellectual infrastructure is not yet mature enough to transform AI into a catalyst for national innovation.

 

Scaling AI development requires immense physical infrastructure: data centers, high-performance chips, advanced digital networks, cooling systems, and an uninterrupted power grid. Indeed, the International Energy Agency estimates that global data center electricity consumption will more than double, reaching approximately 945 terawatt-hours by 2030. Without domestic energy and computing capacity, Indonesian companies will remain dependent on AI services from global companies (IEA, 2025; Ministry of Energy and Mineral Resources, 2025).

 

The government should supervise technology companies involved in AI development. AI impacts citizens’ access to information, employment, credit, education, and public services. If the models and infrastructure are imported, the government has reduced ability to analyze or influence their operation.

 

Conversely, local users must understand the provenance of data, algorithmic limitations, language support, and the critical decisions that must remain in human hands. AI models trained predominantly on foreign datasets risk being structurally blind to Indonesia’s unique social, cultural, and historical contexts.

 

This leads to a fundamental question: Does artificial intelligence empower Indonesian society to cultivate self-reliant intellectual, creative, and critical capabilities, or will it relegate the nation to perpetual reliance on automated computational authority?

 

Shannon Vallor and Tillmann Vierkant (2024) emphasize that current AI systems cannot yet be moral agents. Because these systems lack the moral consciousness and judgment required to be held accountable. Responsibility remains with the humans who design, use, oversee, and regulate them. This is referred to as the “vulnerability gap”: citizens bear the consequences of decisions, but the systems and the people behind them are not in a direct accountability relationship with them.

 

Seth Lazar (2024) highlights AI and its relationship to issues of power. He refers to automated systems that determine what a person can know, obtain, and choose as “automatic authorities.” The use of AI cannot be evaluated solely based on its intended purpose. The manner in which power is exercised and the parties authorized to exercise it must also be scrutinized.

 

If AI models, data, standards, and usage limits continue to be determined overseas, Indonesia will remain merely a user. Being prepared means being able to develop the technology, oversee its implementation, and determine the interests it should serve.