The short version: the Philippines now has a costed national plan for AI infrastructure. The Philippine AI+ Infrastructure Masterplan — PAIIM, covering 2026 to 2033 — was formally launched on September 8 at the Crowne Plaza Manila Galleria in Quezon City. Developed by the Department of Information and Communications Technology (DICT) with the Asian Development Bank and approved by President Marcos as chair of the Economy and Development Council, it puts a number on the country's AI ambition: $34.4 billion (around ₱2.15 trillion) in combined public and private investment over eight years, Newsbytes.PH reported.
"This is our roadmap to make the Philippines more AI-ready — from stronger connectivity, AI computing and data centers, sustainable energy, and digital talent including policies that will help grow the AI ecosystem in the country," the DICT said at the launch.
What PAIIM actually targets
- Grow AI data-center capacity from roughly 50 megawatts today to 1.5 gigawatts by 2033 — about a 30-fold expansion
- Deploy on the order of 152,000 GPUs (the specialized chips that train and run AI models)
- Build a flagship AI compute hub estimated at $14.65 billion (around ₱916 billion): 400MW in phase one by 2030, then 1,100MW more by 2033
- Train more than 1.3 million AI-related professionals
- Create over 500,000 AI-related jobs, plus 175,000 jobs from the infrastructure projects themselves
- Lift national GDP by 10 to 12 percent by 2033
The plan is organized around six focus areas: digital connectivity, AI computing and data centers, energy security, workforce development, policy and regulation, and creating demand for AI services.
Who pays for it
Of the $34.4-billion requirement, about $18.2 billion (around ₱1.14 trillion) is already committed through existing programs, with roughly $16.2 billion (around ₱1.01 trillion) still to be mobilized from public investment and private participation, The Register's analysis notes. The flagship compute hub alone splits into $5.48 billion public and $9.17 billion private money. The private sector is expected to carry the larger share of the total over the plan's life.
Can the grid keep up?
The plan itself flags the hard parts. Launch coverage lists power-grid capacity risks as early as 2027, transmission bottlenecks, GPU scarcity, and even water-treatment limits for data-center cooling. Those constraints are not hypothetical: a 1.5GW data-center fleet would draw more power than some Philippine regions consume today, and every megawatt has to be generated, transmitted, and cooled.
The regional gap is also real. The Register points out that Malaysia's Johor state alone has about 4GW of data-center capacity built or in the pipeline — more than double the Philippines' 2033 target — and the country is competing with Singapore and Thailand for the same investments.
The BPO question
PAIIM addresses the elephant in the room for Philippine tech employment. The masterplan states that it "does not position AI as a replacement for the Philippine IT-BPM industry," arguing instead that AI should handle "first-level, repetitive, and data-intensive interactions" while Filipino workers move up to complex, higher-value customer work. The Register is openly skeptical of that framing, noting the outsourcing industry grew precisely because automation alone never fully replaced human customer service. Either way, the plan ties the country's largest white-collar employer directly to its AI bet.
Why this plan, why now
PAIIM reads as the government's structural answer to a painful recent pattern: hyperscale AI investments landing elsewhere in the region, most visibly when OpenAI anchored two Nvidia-powered AI factories in Malaysia while the Philippines was still mapping its options. It also supersedes months of piecemeal signals about where Philippine AI data centers would actually be built with a single costed roadmap. For Filipino readers, the near-term markers to watch are concrete: whether the 400MW first phase of the compute hub breaks ground on schedule before 2030, and whether the promised training pipeline — 1.3 million professionals — shows up in real course seats and hiring.
