How the service industry sees AI as a double-edged sword

The AI boom has awakened the Philippine service industry to the need to move toward higher-value work. But what will it take to get there?

Service work, including BPO, is knowledge work. It spans call center agents, accountants, doctors, lawyers, bankers and engineers. Like manufacturing, it ranges from low-value execution to highly specialized design, engineering and coordination.

Lower value work

Today, the Philippines remains concentrated in relatively lower-value knowledge work. Much of the highest-value knowledge work stays in the West, where companies research, design and coordinate complex products. Lower-value execution is outsourced to countries like the Philippines.

This encouraged the local service economy to specialize in work that clients had already decided was not worth deeply investing in. These functions were often cost centers, removed from strategic decision-making and given weak incentives for automation. The industry was neither equipped nor motivated to transform them.

The same problem exists within Philippine companies. Because labor is relatively inexpensive, companies often rely on growing headcount instead of standardized processes and software. Manual work accumulates until it becomes operational debt. By the time systems begin breaking under scale, the problem is expensive and difficult to untangle.

So what will enable the Philippine BPO industry to move up the value chain?

The answer is institutional knowledge.

What AI cannot replace

High-value knowledge work emerges when companies develop a deep understanding of a problem, together with the processes, culture and organization needed to solve it repeatedly. That knowledge is built through execution, coordination and iteration.

Our outsourced industries struggle to develop it because few local companies have created environments where it can accumulate. Repetitive manual work persists, while management often substitutes inexpensive headcount for process discipline, automation and enterprise investment.

AI, particularly large language models, is both a threat and a possible antidote to this incentive structure.

As repetitive execution becomes automated, demand and margins for low-value work will fall. At the same time, coding agents are making software faster and cheaper to build. Local companies can increasingly adapt battle-tested, often open source software already widely used in more mature markets.

This can create a healthier cycle. Companies will be pushed toward software, standardization and automation. Work that was once tedious and error-prone will become routine because the solution has been implemented, documented, and improved. These systems will generate institutional knowledge that compounds over time, raising labor productivity and justifying further investment.

This is why focusing only on English proficiency, AI literacy, or critical thinking is insufficient. It treats workers as better raw inputs that can be plugged into higher-value companies. 

But LLMs are rapidly commoditizing many of these capabilities.

The hardest problems in implementing AI are context, coordination and accountability. These are precisely the capabilities the Philippine service industry can develop. These come from deep experience in specific industries, not generic classroom training. The correct answer is often knowable only through the high-resolution context of real work.

Moving up the value chain will be difficult, risky and slow. But building institutional knowledge through automation, standardization and repeated execution is the most realistic path forward.