As the West pushes back on AI, Asia risks becoming its environmental frontier

An anti-AI movement is gaining momentum in the United States, which hosts the world's largest number of data centers, more than 4,400 as of this year, roughly 45 percent of the global total. Communities have been organizing against these facilities for months, and it culminated on July 18 in the first coordinated nationwide effort where 142 protests were held across 42 states. New York became the first state to impose a data center moratorium this year, and polling shows roughly seventy percent of Americans oppose having one built in their own community.

The protests are linked to real harms including massive water consumption for cooling, heavy strain on local electricity grids, and continued reliance on fossil fuels to meet the new demand.

While the protests surge in the US, countries across Asia are scrambling to attract the very investment America is starting to push back on. The flashpoint is India, where the government is courting billions of dollars in AI infrastructure even though it is one of the world's most water-stressed countries.

The clearest example is unfolding in Andhra Pradesh, where Google is building its largest data center outside the US, a fifteen-billion-dollar, one-gigawatt campus near Visakhapatnam, in a district that already has the lowest groundwater availability in the state. The environmental clearance doesn't disclose how much water the facility will actually use once running, and Google hasn't answered questions about its consumption plans.

Data center, however, do have a compelling economic case for countries like India. Development costs in India are among the lowest in the world, second only to China, while electricity prices remain a fraction of what companies pay in the United States, the United Kingdom, or Japan. These advantages help explain why governments are competing so aggressively for AI infrastructure despite mounting environmental concerns. For many policymakers, the promise of billions in investment outweighs warnings about water stress, energy demand, and land use.

Yet the experience elsewhere in Asia suggests that attracting these investments comes with trade-offs that become harder to ignore over time. Singapore offers perhaps the clearest example. In 2019, Singapore imposed a moratorium on new data centers because of concerns over land, water, and energy constraints. When it reopened the sector a year later, approvals were limited to facilities meeting stringent efficiency standards.

Singapore's experience also illustrates another dynamic. Wealthier economies have greater capacity to impose environmental limits on AI infrastructure, but those limits often shift investment rather than reduce it.

The restrictions did not slow regional expansion. Instead, they redirected it. Johor, the Malaysian state bordering Singapore, absorbed much of the investment Singapore declined. It has been the single biggest winner of Malaysia's data center boom since 2022, pulling in about US$39 billion in approved investment by mid-2025  and turning it into Southeast Asia's fastest-growing data center hub. Residents there pushed back this February, taking to the streets over pollution and dwindling water supplies in what became the country's first such demonstration against the industry.

Indonesia is now experiencing the same pattern. In Batam, which already struggles with chronic water shortages, residents protested in December 2024 after data centers kept drawing water while their own homes went dry, and that anxiety over who gets prioritized in the next drought has only grown since. Research estimates that existing and planned facilities could eventually consume around 8 percent of the city's water supply, underscoring how the costs of AI infrastructure are increasingly being borne by local communities.

The Singapore-Johor-Batam corridor offers a preview of what may lie ahead elsewhere in Asia: environmental constraints in one jurisdiction do not slow AI expansion, they simply relocate it to places with weaker regulation and greater economic urgency.

As movements take root in the US and the West and stricter regulations get enforced, the AI industry will eventually look to the Global South as its new frontier, drawn by lax regulations, cheaper electricity, lower costs of development, less organized resistance from environmental activists, and governments hungry for investment. Just as Asia became the hub for producing the world's textiles and fast fashion, AI too is on a path to being outsourced. But the comparison only goes so far. Unlike textiles, AI produces only a fraction of the employment, generates no meaningful export value for host economies, and carries far greater environmental and energy costs.

The difference between the movements taking shape in the Global North and those emerging in the Global South, especially in Asia, is that the latter remain scattered and lack the weight to force real change. The deciding factor is government appetite for investment. Officials are so eager to secure these deals that they ignore or, in some cases, actively suppress local resistance, rushing compliance checks through to expedite projects while disregarding the human and environmental costs.

The choice rests with both governments and people. For governments, it means introducing genuine regulations and compliance mechanisms that account for the human and environmental costs of this infrastructure, not rubber-stamping deals in the name of investment. For communities, getting organized and sustaining protest movements remains the only real leverage available.

Asia faces a difficult choice. It can ride the AI boom and take in billions in investment while watching its lands run dry and its people go without power and water, or it can put the welfare of its people and land ahead of profit.

-Asia Media Centre

Banner Image - Anti-data-center sentiment in rural Kansas neighborhood - Image Credits - Wikimedia