Pax Silica is top of mind right now—impossible to escape in the headlines, the protests, the discussions, the arguments online. Before anyone signs anything, it is worth asking what, exactly, we would be agreeing to.
The pitch is familiar: quality jobs, major investment, a seat at the table in the global AI and semiconductor value chain. It is also not new. The Philippines has heard this kind of pitch before—with Intel, Hanjin, Shell, and Chevron—deals framed the same way, as engines of growth and modernization. Decades later, foreign investors profited handsomely while domestic industry never quite caught up, and the country’s role stayed limited to supplying raw or intermediate materials while design, manufacturing, and packaging happened elsewhere.
The scale of Pax Silica is what makes the stakes worth a close examination. The proposed hub could draw about 3 gigawatts of power and 39 billion liters of water a year. Meeting that demand—including the solar farms and reservoirs needed to supply it—could require a land footprint exceeding 20,000 hectares, land that today could otherwise feed a quarter to half a million Filipinos annually. This isn’t a case against renewable energy or the infrastructure it needs; it’s a reminder that even good tools require honest accounting of trade-offs at this scale.
Who pays and who profits?
That accounting is the pattern worth watching for. The language of progress and opportunity has earlier preceded deals in which the Philippines ended up absorbing the costs while the greater share of the benefits moved elsewhere. Asking who pays and who profits isn’t cynicism—it’s the due diligence that history says we owe ourselves. Until those questions are answered, precaution isn’t the obstacle to development. It is development done properly.
As it stands, Pax Silica risks becoming a deal where the Philippines absorbs the environmental, fiscal, legal, and governance costs, while the benefits—the computing power, the profit, the strategic control—flow outward. That is not a reason to abandon digital infrastructure altogether. It is a reason to demand a different kind of infrastructure, designed around who owns it and what it is built for, rather than around speed and cost alone.
On ownership, the pattern is already global and concentrated. A handful of hyperscale operators—Amazon, Microsoft, Google, Meta—alongside colocation giants like Equinix and Digital Realty, control the overwhelming majority of the world’s data center capacity. They build wherever is cheapest and fastest; the water and electricity are drawn locally, while profit and decision-making stay elsewhere.
On use, the same computing power that could serve hospitals, farmers, and disaster response is instead spent disproportionately on advertising and low-value content, including replacing BPO (business process outsourcing) workers with chatbots. Neither problem is solved by building more data centers faster. Both require deciding—deliberately—who this infrastructure is for.
None of this is an argument against AI. It is not a choice between embracing the technology on its current terms or refusing it altogether. AI is already part of how the world functions, and opting out doesn’t protect a country—it just means giving up a say in how the technology develops. The real choice is narrower: whether AI infrastructure in the Philippines gets built to serve concentrated private profit, or to serve the public. Technology itself is neutral on that question. Design is not.
Smaller, distributed, publicly accountable
The Philippines can pursue smaller, distributed, publicly accountable digital infrastructure instead of one enormous, gigawatt-hungry hub. Picture a network of renewable-powered and water-light facilities located in universities, public hospitals, local governments, and agricultural research centers. Each node could match local demand and use available indigenous renewable energy, with cooling systems designed to conserve water.
In Central Luzon, for instance, a facility could run lightweight models trained on local soil, weather, and planting data to produce crop and rainfall forecasts in Kapampangan or Tagalog—information that municipal agricultural workers can use directly. A public hospital node could process diagnostic tools on-site, from reading X-rays to flagging patient-record anomalies, while keeping encrypted health data local. Disaster-risk agencies could use similarly purpose-built infrastructure to process rainfall, sensor, and satellite data during typhoons, owned by the institutions that respond when disaster strikes rather than rented from a distant platform.
This is what technology for social good looks like when treated as an intentional design choice rather than an afterthought. By favoring smaller physical footprints, public ownership, and community-level governance, we shift the core incentive of digital infrastructure away from maximizing ad-engagement metrics or driving corporate profits.
Digital literacy is also a required part of the infrastructure. Filipinos need to understand how models are trained, where data lives, who owns the servers, how much water and power computation consumes, and who bears the cost when it goes wrong. This is literacy as critical thinking and productive capacity, not passive consumption; it is the foundation of genuine sovereignty over decisions currently being made with minimal public scrutiny.
The guardrails for this transition already exist. The United Nations Secretary-General’s Panel on Critical Energy Transition Minerals has laid out principles—human rights protections, environmental limits, transparency, and fair benefit-sharing—that can and should be applied to the development of all kinds of digital infrastructure.
A sovereign approach is not an argument against ambition or development. It is an argument for building AI infrastructure that the country can stand behind: independently verified, publicly accountable, and designed to serve Filipinos first. By insisting on autonomous design, localized ownership, and unyielding public accountability today, we can ensure that we remain the architects of our own future. CS
Chuck Baclagon is 350 Pilipinas’ senior advisor for campaign strategy and communications, with over 20 years of experience at the intersection of technology and the environment. He co-authored “Pax Silica At Clark: The True Cost Of The AI Rush And A Better Way Forward,” a policy paper published jointly by the Community Legal Help and Public Interest Centre and 350 Pilipinas.
