For the better part of two decades, I’ve watched how governments and industries across Asia have talked about technology adoption. Whenever artificial intelligence (AI) comes up, the default response is to call for more training, more bootcamps, and more university programs. Upskilling, we are told, will secure our place in the AI future.
The Philippines has leaned heavily into this narrative. From coding courses in senior high schools to AI-focused workshops for civil servants, there’s no shortage of rhetoric about building human capital. And to be fair, people are at the heart of any technological revolution. But here’s the uncomfortable truth: skills alone won’t bring AI to the Philippines. We need power.
AI isn’t just about algorithms and clever programmers. It is a profoundly infrastructure-driven technology. Data centers, energy grids, fiber cables, and cooling systems are the true engines of AI adoption. If we train thousands of AI engineers but leave them with no domestic compute infrastructure, we’ve essentially prepared them to work for someone else’s economy.
The infrastructure blind spot
Let’s take a step back. Training modern AI models is resource-intensive. Running them at scale, from large language models to image recognition systems in hospitals, requires high-performance compute (HPC) clusters, cloud regions, and industrial-scale data centers. These in turn demand not just capital but cheap, reliable electricity and resilient networks.
This is where the Philippines stumbles. According to recent regional comparisons, our electricity prices are among the highest in Southeast Asia, roughly $0.20/kWh. By contrast, Malaysia and Indonesia hover closer to $0.10–$0.12/kWh, while Vietnam enjoys rates as low as $0.09/kWh. Brunei, heavily subsidized by its oil and gas revenues, is at $0.07–$0.08/kWh.
These differences are not cosmetic. For data center operators, whether global hyperscalers like Google, Microsoft, and Amazon, or local colocation firms, power costs are the single largest line item. A few cents per kilowatt-hour, multiplied across thousands of servers running 24/7, is the difference between a viable business and an unworkable one.
Why neighbors are winning investments
It is no accident, then, that Malaysia and Indonesia are attracting the lion’s share of new hyperscale investments. Google recently announced a $2 billion data center complex in Johor. AWS and Microsoft continue to expand in Jakarta. Thailand, too, has secured multi-billion-dollar commitments.
What do these countries have in common? Yes, lower power costs. But more importantly, they combine targeted subsidies, government incentives, and infrastructure support that make it feasible for data center operators to scale. Subsidies don’t just make household bills cheaper, they reduce industrial tariffs, stabilize supply, and allow operators to sign special bulk contracts.
Vietnam, for example, has long cross-subsidized residential and social groups, keeping average tariffs low enough to entice data center investments. Even Indonesia, with its uneven grid, manages to offer targeted subsidies for lower consumption brackets and negotiates special power deals for hyperscalers.
Meanwhile, the Philippines has partial subsidies, VAT exemptions for “lifeline” consumers, but no broad industrial policy to make power competitive for data centers. The result? Hyperscale players look elsewhere, and local startups are forced to rely on foreign cloud regions.
Singapore’s exception, and lesson
Critics might point to Singapore, which has some of the highest electricity costs in Asia ($0.25–$0.30/kWh) yet remains the region’s undisputed data center hub. But Singapore is the exception that proves the rule. Its draw is not cheap power but unparalleled infrastructure reliability, digital connectivity, and regulatory transparency. Even with tight land and power restrictions, global firms still want a presence there because of its ecosystem stability.
The Philippines, by contrast, cannot offset high electricity prices with world-class reliability. Our grid remains vulnerable to outages, and our connectivity, while improving, still lags behind regional peers.
What “infrastructure for AI” really means
So what does infrastructure for AI adoption entail? Three things stand out:
- Energy, Affordable, reliable, and increasingly renewable electricity. Data centers run 24/7, and hyperscalers now demand contracts that guarantee green power. Without addressing both cost and sustainability, the Philippines risks being locked out of the AI infrastructure map.
- Data Centers and Compute, Local availability of hyperscale facilities or, at minimum, strong colocation clusters with high-performance compute. Without this, our AI engineers end up renting compute from abroad, which raises costs and undermines data sovereignty.
- Connectivity, Robust fiber backbones, 5G rollout, and international subsea cable redundancy. AI is distributed by nature; latency and downtime kill adoption in sectors like telemedicine, fintech, or logistics.
Overlaying all of these is policy and regulatory clarity, incentives for investment, one-stop permitting processes, and transparent rules for cross-border data flows.
Education plus infrastructure
None of this is to dismiss the value of education. We need more data scientists, machine learning engineers, and AI-literate policymakers. But their skills will not take root in a barren infrastructure landscape. If power costs remain among the region’s highest, if no hyperscale facilities operate locally, if connectivity bottlenecks persist, then upskilling programs become little more than export pipelines, training our best talent for jobs abroad.
The Philippines has the human capital, the English-speaking workforce, and the geographic demand to play a significant role in AI. What it lacks is the hard infrastructure to keep that talent productive at home.
A call to balance the conversation
It’s time to expand the narrative of AI adoption beyond classrooms and workshops. Every peso spent on upskilling should be matched with investments in grid modernization, renewable integration, and data center-friendly policies. Subsidies, carefully targeted, could make industrial power competitive without distorting the broader economy. Incentives could attract the first wave of hyperscale players, seeding the ecosystem for local startups and researchers.
If we don’t, the Philippines will remain a consumer of foreign AI innovations rather than a producer. And all the coding bootcamps in the world won’t change that.
Skills matter. But power, literal, electrical power, is what will determine whether AI truly takes root in the Philippines.
Dominic “Doc” Ligot is one of the leading voices in AI in the Philippines. Doc has been extensively cited in local and global media outlets including The Economist, South China Morning Post, Washington Post, and Agence France Presse. His award-winning work has been recognized and published by prestigious organizations such as NASA, Data.org, Digital Public Goods Alliance, the Group on Earth Observations (GEO), the United Nations Development Programme (UNDP), the World Health Organization (WHO), and UNICEF.
If you need guidance or training in maximizing AI for your career or business, reach out to Doc via https://docligot.com.
