When code writes itself, what happens to Philippine software?

The $6.8-billion slice of our IT-BPM (Information Technology and Business Process Management) industry built on writing code for the world faces an existential question — and the clock is not waiting for us to answer it.

Thirty percent. That is how much of Microsoft’s code is now written by artificial intelligence (AI), according to the company’s own disclosure. Google reports a similar figure, over a quarter. Gartner forecasts that by the end of 2026, 60% of all new code produced globally will be AI-generated. These aren’t projections from breathless futurists. They are operational realities reported by the people signing the paychecks.

For an economy that has bet a significant portion of its future on writing other people’s software, those numbers should keep policymakers up at night.

The Philippine IT-BPM sector closed 2025 with $40 billion in export revenues and 1.9 million workers, according to the IT and Business Process Association of the Philippines (IBPAP). That is more than 8% of GDP. By almost any measure, a success story. But look closer. An AMRO (ASEAN+3 Macroeconomic Research Office) analysis from December 2025 found that contact centers still accounted for 83% of industry revenue. IT and software development services represented just 16% to 18%, or between $6.1 billion and $6.8 billion per ASEAN Briefing. That software slice is the industry’s upmarket frontier, its claim to a future beyond voice calls. It is also the slice most directly in the path of AI coding agents.

The disruption is not theoretical anymore. A Stanford Digital Economy Lab study released in August 2025, drawing on millions of ADP payroll records, found that employment for software developers aged 22 to 25 declined nearly 20% from its late 2022 peak. Older developers held steady or gained. AI is eating entry-level coding work first. Stack Overflow’s 2025 Developer Survey reported that 65% of developers worldwide now use AI coding tools at least weekly. Last February, one engineer at a major San Francisco tech company told the SF Standard that his entire job had become acting as a proxy. His manager tells him what to do, and he tells Claude to do it.

This is where the Philippine exposure gets specific. A February 2025 IMF working paper on the Philippine labor market found that roughly one third of Filipino workers are highly exposed to AI. But it also noted that about 61% of those exposed jobs are “highly complementary” to AI, meaning productivity could rise if workers learn to use the tools rather than compete against them. The operative word is “if.” The IMF’s finding is a conditional promise, not a guarantee. Complementarity only materializes through deliberate investment in skills, tools, and institutional redesign.

The problem is that the Philippine software outsourcing model was built on a specific value proposition: English-speaking developers at 60% to 70% lower cost than American equivalents, producing competent code on a follow-the-sun schedule. That proposition assumed coding labor was the bottleneck. AI coding agents are dissolving it. A senior architect in Taguig and a senior architect in Ho Chi Minh City both become less differentiated when the routine code between their design decisions gets written by a machine in seconds.

And Vietnam is not standing still. In December 2025, Vietnam’s National Assembly passed a comprehensive AI law, the first standalone AI legislation in Southeast Asia, set to take effect this month. The country is producing over 55,000 tech graduates annually, has attracted AI-focused R&D centers from Samsung, Qualcomm, and NVIDIA, and is actively positioning itself as an AI-native development hub rather than a traditional outsourcing destination. The Philippines’ regional competitors aren’t just adapting to the same disruption. They are building legal and institutional frameworks to capture the next wave while we are still debating how to protect the last one.

So, what do we do about it? Three things, with urgency.

First, IBPAP, the Department of Information and Communications Technology (DICT), and the major outsourcing firms need to redefine the Philippine value proposition away from “we write code cheaper” toward “we deliver AI-augmented outcomes faster.” This requires equipping Filipino developers with AI coding tools and training them to orchestrate, validate, and govern AI-generated output. The companies that wait will find their clients replacing offshore teams with smaller domestic teams armed with AI.

Second, the education pipeline needs a hard reset. CS programs at UP Diliman, Ateneo, De La Salle, and the State Universities and Colleges feeding the IT-BPM talent pool cannot keep teaching students to write code the way they did in 2019. The Stanford data is clear: the market for entry-level developers who only know textbook algorithms is collapsing. Filipino graduates need to emerge fluent in AI-assisted development, prompt engineering, system design, and the security oversight of AI-generated code. Recent industry assessments show that 45% of AI-generated output still contains vulnerabilities. The Commission on Higher Education should be coordinating curriculum reform with IBPAP on a crisis timeline.

Third, the IBPAP roadmap targeting $59 billion in revenue and 2.5 million jobs by 2028 needs a credibility check. Those targets were set before AI coding agents reached their current capability. Growth will come from AI-augmented high-value services, not from adding more seats writing more lines of code. The roadmap should be stress-tested against a scenario where AI reduces the labor content of software delivery by 30% to 50% within three years. If that exercise has not happened, it should happen this quarter.

The Philippine IT-BPM industry has survived predictions of its demise before. Chatbots. The offshoring backlash. The pandemic. Each time, it adapted. But those were threats to efficiency. This one is a threat to the core product. When the code writes itself, the country that still sells code-writing labor holds a depreciating asset.

The $40-billion question is whether we depreciate with it. The window for deciding is not five years. It is now.

This article reflects the personal opinion of the author and does not reflect the official stand of the Management Association of the Philippines or MAP.

Itamar Gero is a member of MAP NextGen Committee. He is founder and CEO of Truelogic, Inc., an AI-enabled digital performance agency working with the country’s largest brands on their digital strategies.

[email protected]

[email protected]