Jakarta, INTI - Artificial intelligence (AI) data centers generate significantly more heat than conventional servers typically used for everyday web browsing. As a result, AI-focused data centers require increasing amounts of water to support their cooling systems.
In Indonesia, most data centers still rely on conventional air-cooling systems. These systems depend heavily on water circulation to cool the air and maintain optimal temperatures inside data halls.
AI Data Center Put Pressure on Water Resources
Secretary General of the National Association of Information and Communication Technology Entrepreneurs (APTIKNAS), Fanky Christian, warned that the growing water requirements of AI data centers could raise concerns over competition for water resources with other sectors.
“There will be competition between household needs and data centers. If you own the water supply, which one would you choose to sell to? Data centers offer different pricing because they are industrial-grade customers, while residential customers may even receive subsidized rates,” Fanky said in Jakarta on Friday (August 28, 2026).
The potential for water stress linked to AI data center operations has already emerged in industrial areas such as Johor Bahru, Malaysia. Heavy groundwater consumption by the data center industry has reportedly prompted concerns among nearby residents over access to water resources.
Cooling capacity is also becoming a major challenge for the development of Indonesia’s AI data center industry. Rising temperatures driven by the climate crisis are adding further pressure to cooling systems and their capacity requirements.
In Jakarta, several data center facilities are being developed in urban areas located near lakes or rivers, including parts of Kuningan and MT Haryono, to support the sustainability of their cooling operations.
Liquid Cooling Emerges as a More Sustainable Solution
As a long-term solution, Fanky suggested that AI data center operators should increasingly consider liquid cooling technology.
According to Fanky, liquid cooling can be more environmentally friendly than conventional air cooling. Efforts are currently underway to promote cooling-system standards, including the adoption of liquid cooling for AI data center infrastructure.
“From an environmental perspective, the impact could be much better because the system does not rely entirely on environmental water resources. However, liquid cooling is used specifically for the equipment, while the facility itself still requires water because the chiller system needs to circulate, cool, and discharge water,” he explained.
Conclusion
The rapid expansion of AI data centers is creating new challenges for Indonesia, particularly in managing water resources and cooling capacity. As demand for AI computing continues to grow, adopting more efficient technologies such as liquid cooling could help data center operators balance performance, resource efficiency, and environmental sustainability.
Read more: Rosan Calls for Investment in AI-Ready Infrastructure to Position Jakarta as a Data Center Hub