In February 2025, an IMF working paper named business process outsourcing as the Philippine sector with the highest proportion of jobs at risk of displacement from AI. Roughly eighteen months later, the contact center industry reported it had added more than 60,000 workers. Both facts are true. The distance between them is where the useful information lives.

This analysis pulls together the three data sets that matter most for anyone planning a workforce or a career: the IMF's occupational exposure study, the ILO's 2026 generative AI briefs, and the industry's own employment figures. Read together, they give a more precise answer than either the doom headlines or the reassurance campaigns.

What the IMF study actually found

IMF Working Paper WP/2025/043, by Micholo Cucio and Tristan Hennig, combined Philippine labor force survey microdata with measures of AI exposure and complementarity. The headline result: around one third of Filipino workers are highly exposed to AI, and around sixty percent of those are also rated highly complementary, meaning the technology is more likely to assist them than replace them.

The figure that traveled furthest is the sector cut: 89 percent of the BPO workforce is classified as highly exposed to AI. Quoted on its own, that number reads like a countdown. It is not. In the same analysis, 61 percent of those highly exposed BPO jobs are rated highly complementary. Exposure, in this methodology, measures whether AI can perform a meaningful share of an occupation's tasks. It says nothing about whether an employer will cut the job. A role can be 89 percent exposed and still get more valuable, if the human moves up to the work the model cannot do.

Coverage of the IMF findings in the Daily Tribune puts the displacement side in sharper focus: about 14 percent of Philippine jobs face a high risk of displacement, while 22 percent are expected to undergo significant transformation. Displacement and transformation are different outcomes, and the second group is larger.

The ILO data points to task change, not mass layoffs

The ILO's February 2026 country brief measured generative AI specifically. It found that more than a quarter of Philippine employment, about 12.7 million workers, is exposed to generative AI, the highest rate among comparable ASEAN countries. The same brief found that only 3.6 percent of total employment sits in the highest exposure category where displacement risk is most acute. The ILO's own reading, as reported by BusinessMirror, is that the impact will come "through task transformation within occupations rather than full job replacement."

Its July 2026 regional follow-up added a reality check on timing. The Philippines ranks second in ASEAN for exposure, with 28.1 percent of employment in occupations with more than minimal GenAI exposure, behind only Singapore at 42.2 percent. Yet the ILO reported that "widespread disruption is not yet visible" and that jobs in highly exposed occupations were still growing across the region. The warning buried in that brief is different from the one in the headlines: the Philippines ranks second in exposure but only fourth in preparedness. The gap between those two rankings, not the exposure number itself, is the risk.

Employment is still rising, but the ceiling dropped

The industry's own numbers show growth, not contraction. The Contact Center Association of the Philippines reported that the sector added more than 60,000 workers in 2025, lifting full-time contact center employment to 1.68 million, with 1.73 million expected in 2026. Industry-wide, IBPAP counts 1.9 million IT-BPM workers and over $40 billion in revenue for 2025, and projects $42.3 billion and 1.96 million workers for 2026.

The honest caveat came in July 2026, when IBPAP cut its 2028 roadmap targets. The original 2022 roadmap promised $59 billion in revenue and 2.5 million jobs by 2028. The revised range is $43.3 billion to $50.5 billion, and 1.85 million to 2.14 million workers, with AI adoption named among the reasons. Note what the downside scenario implies: 1.85 million in 2028 would be fewer workers than the industry employs today. That is a real possibility on the table, stated by the industry's own association.

"I think AI is a real development, but I think we have not really seen it scale yet. It has affected some jobs, but based on all the members we've talked to, the entry-level jobs that some of the AI trials have affected, those employees have been able to be redeployed." Jack Madrid, IBPAP president and CEO, July 2026

Early company-level evidence matches that measured tone. Fortune's reporting on the sector found that while 56 percent of back-office firms were incorporating AI tools, only 8 percent had cut headcount because of automation as of early 2025.

The risk concentrates in specific roles, and so does the growth

Averages hide the real story. On the displacement side, the ILO found that clerical support workers face the highest exposure, 3.5 million jobs, with 37.8 percent in the highest displacement-risk category. Exposure is also unevenly distributed by gender: 40.3 percent of jobs held by women show high GenAI exposure versus 19.3 percent for men, and in Metro Manila around two in five jobs are exposed. In practice, this is the territory where contact center AI platforms are aimed first: scripted, high-volume, low-complexity interactions. An AI customer service agent handles the password reset and the balance inquiry long before it handles the irate customer with a five-year account history.

On the growth side, the same industry that is automating transactional work is hiring for roles that did not exist a few years ago. CXAP leadership named the categories directly in May 2026:

For scale on the displacement side, one Manila Times analysis estimated that roughly 57,000 BPO workers face direct displacement risk, derived by applying a 3 percent immediate-risk rate to the 1.9 million workforce. Treat that as an illustrative estimate, not a forecast. It is, however, a very different number from 1.9 million.

What workers and managers should do with this

The data supports three practical conclusions. First, the question "will AI replace call center jobs" has a role-level answer, not a yes or no. Voice and clerical roles built on scripted volume are shrinking territory; judgment-heavy, relationship-heavy, and AI-adjacent roles are expanding. Second, the window for repositioning is open now, while employment is still growing and while employers are funding the transition: IBPAP has committed to at least $25 million per year in workforce development, alongside Project UNLAD, a 740 million peso upskilling initiative with DICT and TESDA. Third, for team leaders and L&D managers, the exposure-versus-preparedness gap the ILO flagged is the actionable metric. Your agents are already exposed; the variable you control is how prepared they are.

If you are mapping specific skills to specific roles, our AI Academy tracks are built around exactly this transition. The honest summary of the evidence as of August 2026: AI is changing call center jobs in the Philippines faster than it is eliminating them, the elimination risk is real but narrow so far, and the workers who move toward the complementary side of the 89 percent will be the ones the next roadmap counts.