From Rented Labor to Owned Capability

What Happens to a Country Built on Rented Labor?

The Philippines built prosperity by renting out its workforce. AI is ending the lease.

I am an American, but my daughter is Filipino. I came to the Philippines not as a consultant passing through, but as someone who chose this country, built a career here, and stayed. When I write about the fragility of what this economy has created, I am not writing from a comfortable distance. I am writing from inside the blast radius.

That is why the question of what comes next is not academic to me. It is personal.

The bet that won

Twenty-five years ago, the Philippines made a smart decision.

The country looked at what it had: a workforce fluent in English, disciplined in service, culturally wired to work with Western clients, and hungry for formal-sector jobs that the domestic economy could not supply at scale. It looked at what the world wanted: talented, affordable people to handle the cognitive work that corporations in the United States and Europe needed done but did not want to pay American or European wages to do.1

The match was obvious. The bet was rational. And it worked.

IT-BPM became a macro-critical pillar of the economy. It created a genuine middle class, stabilized dollar inflows, and gave millions of Filipinos a formal-sector path that much of the domestic economy could not match. By early 2026, the sector was tracking toward $42 billion in export revenues and nearly 1.97 million jobs. It surpassed overseas remittances as the country’s top dollar earner and became approximately 8 percent of GDP.23

This was not fake growth. Millions of Filipino families moved forward because of that bet. The dignity in that is real, and it deserves to be named.

But here is the part that rarely gets told at the industry galas and the government press releases.

The Philippines won the race it entered. It just did not notice that the race had changed.

The bargain underneath the bargain

Too much of the local debate still swings between two lazy positions. One says BPO is a triumph and therefore must remain the backbone of the economy. The other says BPO was always hollow and therefore deserves no defense. Both are wrong. The industry is real, large, and valuable — but also narrow, externally controlled, and increasingly exposed to the exact category of technological change now moving fastest.43

The offshore bargain was simple. Foreign firms brought clients, brands, processes, and global contracts. The Philippines provided disciplined labor, tax incentives, office parks, and scale. As long as labor arbitrage remained durable, the arrangement looked like a stable development model.

But what looked stable was, in reality, conditional. It depended on one thing above all: that the cheapest way to handle a large share of routine cognitive work would continue to be hiring Filipino workers by the thousand.43

That made it a tenancy model. The Philippines was a very good, very reliable tenant. But tenants do not own the building. And when the building gets redesigned from the ground up, tenants are the first to find their lease at risk.

The most important analytical mistake people make about this transition is treating it as mainly a technology story. It is not. It is an ownership story — about who owns the firms, who books the profits, who holds the intellectual property, who makes the strategic decisions, and where the gains from productivity ultimately land.

The Philippines retained real domestic value from the BPO era, especially through wages, office demand, and local operating spend, but it did not build enough Filipino-owned capability underneath that value stream. It captured income without capturing enough control.43

The race that changed

AI is not simply replacing workers. It is renegotiating the entire value of the offshore arrangement.

When the product is mostly human attention, process compliance, and scripted interaction, the key advantage of an offshore location is wage differential. But once models absorb more of the routine work, the value of that wage differential narrows. The relevant questions shift entirely.

Which country owns the compute stack? Which market has the customers? Which firms can redesign workflows end-to-end? Which governments are trying to internalize the gains? Which workforces are being retrained for higher-value layers of the stack rather than simply defending yesterday’s headcount?567

Those are not Philippine strengths — at least not yet.

Compounding this is a political shift in the major customer markets. Once AI begins threatening white-collar labor in the United States and Europe, the old offshore bargain stops making sense for them too. Firms have more reason to internalize margin and control. Governments have more reason to frame the transition as competitiveness, sovereignty, and protecting domestic workers. And all of this is unfolding in a political climate already more comfortable than it was a decade ago with economic nationalism, industrial policy, and onshoring.56789

The Philippine danger, then, is not simply unemployment. It is that the country can look more productive on paper while becoming more fragile underneath — in a world where the major customer markets are becoming more willing to pull value back inside their own borders.

The India mirror

Consider the comparison that reveals the stakes most clearly: India.

India is being hit by the same AI wave. Reuters has already documented routine customer-service automation, slower hiring in BPM, and real staffing pressure as AI agents take over more repetitive tasks. The CEO of Tata Consultancy Services — India’s largest IT firm — told the Financial Times that AI could reduce the need for call centers to “minimal” levels.1011

So the comparison is not between one country facing disruption and another escaping it. The comparison is between two countries facing disruption from very different structural positions.

India enters the transition with large domestic champions, deeper capital markets, a much stronger history of retaining profit pools at home, and more explicit national capability-building. The IndiaAI Mission carries a $1.25 billion government budget over five years, with support for indigenous foundational models trained on Indian data, over 38,000 GPUs of publicly backed compute capacity, and 1.6 million workers already enrolled in a joint government-industry digital reskilling program.12

That does not eliminate pain. But it gives India more levers. It has firms that can pivot from labor-intensive delivery toward higher-value AI-enabled services without the gains automatically leaking abroad. It has more institutional capacity to turn disruption into domestic capability.

India is exposed too. The difference is not exposure. The difference is agency.1012

The Philippines, by contrast, still looks too much like a tenancy model. It hosts globally integrated delivery operations. It benefits from wages and local spend. But too much of the strategy, profit allocation, and technology ownership lives elsewhere. That was tolerable when the country’s role in the value chain was hard to substitute. It becomes dangerous when the value chain is being redesigned around AI.43

The numbers nobody puts next to each other

The scenario modeling I built for this analysis makes the exposure concrete.

Using an event-tree framework calibrated to Philippine conditions — where labor-cost recycling capacity is lower and financial amplification risk is higher than in the US — the probability-weighted analysis suggests the systemic crisis tail is material: crisis-level scenarios carry roughly a one-in-six probability, while the broader recession-and-displacement bucket approaches one-in-three.13

In scenarios where AI adoption is broad and automation penetration is high, total Philippine GDP still rises — roughly 1.19x to 1.29x baseline over five to ten years — but total employment falls to roughly 0.83x to 0.88x, and the current account deteriorates from –3.4 percent of GDP to as low as roughly –7.5 percent as BPO export demand weakens.131011

These ranges are consistent with the IMF’s February 2025 finding that about 14 percent of the Philippine workforce sits in high-exposure, low-complementarity roles, and with the ILO’s assessment that 89 percent of Philippine BPO workers face high automation risk.

In other words: AI can make the Philippine economy look stronger in aggregate while making Philippine society less resilient in practice.

The industry’s concentration makes the exposure worse. BPO remains the majority of sector revenue and roughly 88 to 90 percent of employment. Contact centers still dominate the job base. That concentration was once a strength because it enabled scale. Under AI diffusion, it becomes a vulnerability because routine customer-service and back-office workflows are precisely where substitution, augmentation, and pricing pressure arrive first.43

How we got here

The structure around that work makes the exposure worse still. The country offered highly generous tax holidays and special regimes to attract foreign operators, but it did not pair those incentives with a serious long-term strategy to build local champions, deepen domestic technology ownership, or create meaningful employment-linked clawbacks if automation later reduced headcount.

Government perfected hospitality as industrial policy. It became very good at welcoming other people’s companies into PEZA zones and very poor at ensuring that a large enough share of the strategic upside would remain Filipino when the model changed.43

Industry is not blameless either. For years, the warning signs were visible: task concentration in voice and low-complexity service work, flat career ladders, heavy dependence on North American demand, growing importance of automation in client roadmaps, and a local ecosystem that was much stronger at producing managers for delivery centers than founders of globally competitive technology firms. The industry spent decades proving the Philippines could execute for foreign clients. It spent far less energy proving the Philippines could own more of the stack.43

Filipino capital, too, was often more comfortable financing property, consumption, and service expansion than patient bets on domestic technology capability. That choice had logic: office towers, townships, and service ecosystems around BPO cash flows were highly profitable while the model was ascending. But it also meant that the domestic spillover attached to rented labor rather than owned intellectual property. The country got payroll, rent, and urban consumption. It got far less compounding control.43

The blast radius

At this point, a reasonable person might say: fine, BPO faces pressure, but surely the impact stays in one industry.

It does not.

IT-BPM accounted for 44 percent of all Philippine office transactions in 2024. The sector now accounts for approximately 8 percent of GDP and a comparable share of foreign exchange inflows. A large BPO contraction would not stay inside BPO. It would run through commercial real estate first (vacancy rates, developer revenues, the REIT sector) and from there through urban consumption, household credit, transport, retail, and the foreign-exchange position.314

Manila’s office districts are, in a real sense, a physical expression of the BPO bet. The gleaming towers of BGC and Ortigas exist because of that bet. The malls, the condominiums, the informal service economy that grew around all of it — these are downstream of the same wager. A structural contraction in the sector is not an industry problem. It is an urban problem, a household problem, and eventually a fiscal problem.

Even without a nominal GDP collapse, the stress testing underlying this piece suggests the current account could deteriorate by two to four percentage points of GDP under broad-automation scenarios. That kind of shift — gradual, not sudden — can still create a visibly more brittle economy. One that looks fine in the aggregate statistics while quietly becoming harder for ordinary families to navigate.

The blast radius is not theoretical. It is built into the way the Philippine growth model has been structured for a generation.31314

The gap that should make you angry

This is also why the response cannot be a sentimental defense of the old BPO model. That model still matters. It still supports millions of people. It still buys time. But buying time is not the same as solving the problem. The soft landing is not preserving legacy headcount forever — it is using the current BPO base as a bridge into something more locally owned and AI-native before the bridge decays.23

The country already has the beginnings of policy movement. The National AI Strategy was approved in 2025. The DOST-PCIEERD roadmap points to AI data centers in selected higher educational institutions, AI virtual hubs, and mission-driven AI projects. Something is moving, but not yet at the scale the moment demands.153

How far from that scale? The numbers are clarifying.

The government’s flagship AI reskilling program for IT-BPM — Project UNLAD, a joint TESDA-DICT-IBPAP initiative — carries a four-year budget of ₱740 million. Roughly $13 million. For a 1.9-million-strong workforce, that works out to about $1.70 per worker per year in dedicated government transition funding.16

IBPAP member companies collectively spend roughly ₱1.4 billion a year across learning and development programs. But corporate L&D budgets are not a national transition strategy. They optimize workers for today’s delivery model; they do not prepare an industry for a structural break.17

Even combining all projected government-funded AI training for 2026, coverage reaches approximately 68,000 learners — 3.6 percent of the workforce in a single year.1618

Now put that next to Thailand — a country in the same region with a smaller population and comparable GDP — which committed roughly $714 million to government AI programs over just two fiscal years in mid-2025, including dedicated funding for workforce development, centres of excellence, and a national data bank. On a per-capita basis, Thailand is spending roughly $9.90 on government AI for every $0.38 the Philippines spends.

That’s a 26-fold gap between two countries at roughly the same development stage.19

The Philippines is not spending nothing. But it is trying to navigate a generational industry transition with a rounding error for a budget.

What the country already has

Here is where the story turns. Because the Philippines has more raw material for this transition than the pessimists admit.

This country has deep domain expertise in customer operations, compliance-heavy workflows, training, workforce management, finance and accounting processes, healthcare support, and multilingual service environments. It has a large base of young, adaptive workers. It has diaspora links. It has an English-language advantage that took decades to build and cannot be easily replicated. These are genuine assets — the kind of assets that, in the right hands, could be converted into something far more durable than labor hour-arbitrage.

The window is still open. But it will not stay open indefinitely.

What is needed is a Filipino entrepreneurial renaissance — not as branding, but as economic necessity. Founders need to stop seeing AI primarily as a tool that threatens local jobs and start seeing it as the cheapest opportunity in decades to convert deep service knowledge into owned products, workflows, and outcome-based firms. Investors need to stop treating domestic technology as peripheral to the real economy and start recognizing that local capability formation is now a macro hedge. Enterprise executives need to stop imagining that transformation means merely buying foreign AI tools more cheaply; it also means building local implementation capability and new revenue models.

Mid-career workers, especially those in the upper layers of BPO operations, are perhaps the most important constituency of all. They are not liabilities waiting to be automated away. They are the founding class of the next wave, if they are given the tools, the capital, and the cultural permission to see themselves that way.

What has to change

The next model has to move from labor-hour arbitrage to outcome-based delivery, from foreign-controlled profit pools to more Filipino-owned firms, and from passive upskilling rhetoric to actual grassroots capability formation. The Philippines needs more deployment shops, more service-to-software wedges, more vertical AI products, more local demand creation, more founder formation, and more institutions that turn workers into builders rather than waiting for displaced workers to absorb downward mobility as if it were policy.

That will require a harder politics than the country is used to.

Government should be pressed on incentives, accountability, and transition planning that actually matches the scale of the moment. Industry should be pressed on whether it wants to be remembered as the coalition that maximized headcount during the boom or the coalition that helped build local capability before the bargain broke. Local investors should be pressed on whether they want to keep underwriting secondary spillovers from foreign-controlled sectors or begin financing institutions that can actually own a larger share of the next value chain.

And all of this needs to happen quickly enough that the window remains open.

From rented labor to owned capability

For twenty-five years, the Philippines proved it could be indispensable to other people’s companies.

The next five years will decide whether it can become more indispensable to itself.

The country built a middle class on rented labor. There was dignity in that, and real value in it. But the next middle class will have to be built on owned capability.

My daughter will live in whatever future the country chooses to build. That is why I am writing this now, while there is still time to change it.

Footnotes

  1. OECD, Economic Surveys: Philippines 2026, February 12, 2026. https://www.oecd.org/en/publications/oecd-economic-surveys-philippines-2026_f0e0c581-en.html ↩

  2. “IT-BPM sector on track to hit $40 billion revenue goal,” Philstar, January 1, 2026. https://www.philstar.com/business/2026/01/01/2498004/it-bpm-sector-track-hit-40-billion-revenue-goal ↩ ↩2

  3. ASEAN+3 Macroeconomic Research Office (AMRO), “Can the Philippines’ IT-BPM Industry Stay Ahead Amid the AI Wave?” 2024. https://amro-asia.org/can-the-philippines-it-bpm-industry-stay-ahead-amid-the-ai-wave/ ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13

  4. World Trade Organization, Trade in Services for Development: Case Study — Philippines, 2025. https://www.wto.org/english/tratop_e/ts4d_e/case_studies_e/philippines.pdf ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8

  5. James van Geelen and Alap Shah, “The 2028 Global Intelligence Crisis,” Citrini Research, February 22, 2026. https://www.citriniresearch.com/p/2028gic ↩ ↩2

  6. Executive Order 14179, “Removing Barriers to American Leadership in Artificial Intelligence,” The White House, January 23, 2025. https://www.whitehouse.gov/presidential-actions/2025/01/removing-barriers-to-american-leadership-in-artificial-intelligence/ See also: “Winning the Race: America’s AI Action Plan,” July 23, 2025. https://www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdf ↩ ↩2

  7. European Commission, “AI Continent Action Plan,” COM(2025) 165, April 9, 2025. https://digital-strategy.ec.europa.eu/en/library/ai-continent-action-plan ↩ ↩2

  8. “Far right makes gains in EU election,” Reuters/VOA, June 9, 2024. https://www.voanews.com/a/far-right-makes-gains-in-eu-election/7649465.html ↩

  9. International Monetary Fund, World Economic Outlook: Global Economy in Flux, Prospects Remain Dim, October 2025, Chapter 3. https://www.imf.org/en/publications/weo/issues/2025/10/14/world-economic-outlook-october-2025 ↩

  10. “Meet the AI chatbots replacing India’s call center workers,” Reuters, October 15, 2025. https://www.reuters.com/world/india/meet-ai-chatbots-replacing-indias-call-center-workers-2025-10-15/ ↩ ↩2 ↩3

  11. “AI could kill off most call centres, says Tata Consultancy Services head,” Financial Times, April 25, 2024. https://www.ft.com/content/149681f0-ea71-42b0-b85b-86073354fb73 ↩ ↩2

  12. IndiaAI Mission, Call for Proposals to Build Foundational AI Models, Government of India, 2025. https://indiaai.gov.in/article/indiaai-mission-call-for-proposals-to-build-foundational-ai-models See also: Press Information Bureau, January 30, 2025. https://www.pib.gov.in/PressReleasePage.aspx?PRID=2097709 ↩ ↩2

  13. International Monetary Fund, “Artificial Intelligence and the Philippine Labor Market: Mapping Occupational Exposure and Complementarity,” Working Paper WP/25/43, February 2025. https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025043-print-pdf.pdf ↩ ↩2 ↩3

  14. Colliers Philippines, “IT-BPM Industry Sees Strong Growth and Strategic Shifts in 2024,” BPO Situationer Part 5, early 2025. https://www.colliers.com/en-ph/news/bpo-situationer-part-5-it-bpm-strong-growth-strategic-shift-2024 ↩ ↩2

  15. DOST-PCIEERD AI Roadmap; National AI Strategy for the Philippines (NAIS-PH), approved May 2025. https://pco.gov.ph/news_releases/pbbm-make-best-use-of-ai-for-national-devt/ See also: DOST, April 30, 2025. https://www.dost.gov.ph/knowledge-resources/news/86-2025-news/3989-dost-builds-on-ai-national-strategy.html ↩

  16. Project UNLAD (Uplifting National Labor through AI and Digital Skilling), ₱740 million over four years, implemented by TESDA in partnership with DICT and IBPAP under Republic Act 12063. https://newsbytes.ph/2026/01/30/ibpap-sets-skills-focused-agenda-for-2026-as-ai-reshapes-it-bpm-industry/ Projected 2026 coverage of approximately 68,000 learners per Manila Times, February 24, 2026. ↩ ↩2

  17. IBPAP member company training investment of approximately ₱1.4 billion annually across learning and development programs. “Philippines BPOs invest $25Mn in upskilling to combat AI threat,” Outsource Accelerator, 2026. https://news.outsourceaccelerator.com/bpos-25mn-upskilling/ ↩

  18. DOST Secretary Renato Solidum Jr., 1st National AI Stakeholders’ Conference, October 13, 2025: cumulative Philippine government AI investments exceeding ₱2.3 billion across 113 projects since 2017. Reported by Manila Times, October 14, 2025. DOST subsequently announced plans to invest over ₱2.6 billion for AI projects through 2028 (PNA, June 2025). https://www.pna.gov.ph/articles/1251777 ↩

  19. Thailand National AI Committee, government AI budget of ฿25 billion (~$714 million) approved over fiscal years 2026–27, June 2025. https://www.bangkokpost.com/business/general/3078438/b25-billion-approved-for-ai-development ↩