The Philippines Bets $34.4 Billion It Can Build an AI Hub Without Hurting Its Outsourcing Industry

The Philippine government on September 8, 2026, released the final draft of its Philippines AI+ Infrastructure Masterplan (PAIIM), a $34.4 billion, 2026-to-2033 roadmap the Department of Information and Communications Technology (DICT) says could lift national GDP by 10 to 12 percent by 2033 while making the country Southeast Asia’s leading AI hub, according to The Register. DICT frames that GDP figure as an expected outcome rather than a guaranteed one, according to Newsbytes.ph, which also reports the plan was approved by President Ferdinand Marcos Jr. in his capacity as chair of the country’s Economy and Development Council and formally launched at an event in Quezon City.

The numbers set up an obvious tension, one The Register calls out directly: one of the clearest proven uses for generative AI today is automating the customer-service work that outsourcing already exists to handle cheaply, meaning AI could undercut the same low-wage offshoring model that built the Philippines’ outsourcing industry in the first place. PAIIM does not try to dodge that tension. The plan treats the country’s 1.3-million-strong IT-BPM workforce as an asset to retrain rather than a casualty to manage, and leans on that same workforce, plus the country’s location and existing digital infrastructure, as part of the pitch it is making to the hyperscalers it is trying to attract.

What the Money Buys

PAIIM calls for $34.4 billion in combined public and private investment through 2033, with the Philippine government expecting roughly $13.5 billion, or 39 percent, to come from public sources and the remaining $21 billion, or 61 percent, from private investors, according to The Manila Times. Most of that money is aimed at one number: the country’s AI-oriented data center capacity, currently about 50 megawatts spread across a single facility, is supposed to grow 30-fold to 1.5 gigawatts by 2033, with an initial 400-megawatt phase due by 2030. By 2033, the plan expects that capacity to house around 152,000 GPUs, with 40 percent of the power behind it coming from solar and geothermal sources.

DICT frames the goal in its own words: “By leveraging the country’s strategic location, strong talent base, and established IT-BPM sector, the Masterplan aims to elevate the Philippines into Southeast Asia’s premier AI hub. This Masterplan provides a blueprint for turning the Philippines’ AI ambitions into reality,” the agency said, according to BusinessWorld. That outlet also reports the plan already has a running start: at least two U.S. hyperscalers are reportedly planning to begin building 400-megawatt data centers in the country as early as next year.

A Small Bid in a Region Placing Big Ones

Government officials framed the plan as a way to compete with regional rivals Singapore and Thailand, The Register reports, but notably left one country off that list: Malaysia, where the Johor data center corridor already hosts almost 4 gigawatts of capacity, more than double the Philippines’ entire 2033 target, with a similar amount reportedly still in the planning and construction pipeline. Japan, for its part, is separately pursuing about $60 billion in AI data center investment through 2033, according to Nikkei Asia.

Measured against the biggest spenders in the market, the ambition is modest by design. The Register notes that companies like Amazon, Google, and Microsoft currently spend roughly $20 billion on AI infrastructure every quarter, with Chinese cloud providers adding tens of billions more every year, and frames the Philippines’ entire $34.4 billion plan as bidding for less than a month of that combined big-tech spending.

Jobs, Corridors, and a Workforce Bet

PAIIM projects the buildout will create more than 500,000 AI-related jobs by 2033, including 175,000 tied directly to infrastructure construction and operations, on top of retraining the 1.3 million existing IT-BPM workers already employed in the country’s outsourcing sector, according to the Manila Times. DICT’s Gemma Baysic, officer-in-charge of the department’s National ICT Planning, Policy and Standards Bureau, told the outlet the goal is for the Philippines to build, host, power, and support the AI ecosystem itself, rather than simply consume AI technology built elsewhere. Energy Undersecretary Maria Francesca Del Rosario was also involved in outlining the plan’s renewable-energy targets, the Manila Times reported.

Geographically, the plan anchors around the Clark-Bataan corridor as its primary hub, with Batangas-Aurora serving as a secondary gateway backed by Subic and Calabarzon, while Cebu, Iloilo, Davao, and Cagayan de Oro are positioned as future regional nodes once the initial buildout is underway. DICT is betting that the country’s existing assets, including more than 20 submarine cable landings, over 95 percent mobile coverage, and a workforce already fluent in serving Western clients, are enough to make that geography attractive at a fraction of what its largest competitors are spending.

The contrast with some of the Philippines’ neighbors is notable. Just days earlier, Thailand froze all new data center approvals nationwide and paused 166 projects already in the pipeline while it wrote rules for an industry its own officials admitted they had never fully measured. The Philippines is moving in the opposite direction: trying to build an AI data center industry up from a 50-megawatt base rather than rein one in, betting that a coordinated national plan and a workforce built for decades of global outsourcing can make up for arriving late and small to a race the biggest players are running at tens of billions of dollars a quarter.