In Philippines, 16.63% of sampled conversations were matched to software development tasks. In Korea that figure is 10.42%.

What the data says about Philippines

Here is the number that should flip any easy assumption about who uses AI “seriously”: 16.63% of sampled conversations from the Philippines were matched to software development tasks, compared with 10.42% from Korea. That is nearly 60% higher than a country whose overall usage index is more than eight times larger (3.78 against the Philippines’ 0.44). Fewer conversations in total, but a much heavier concentration of code inside them. If you expected intensity of use and technical depth of use to travel together, this country’s data says otherwise.

The profile gets more specific from there. Content creation and copywriting accounts for 28.82% of conversations, well above the 22.72% global share, and conversations matched to arts, design, entertainment, sports, and media job tasks run 18.99% against a global 13.55%. Research and intelligence work sits at just 5.8% — barely half the global 10.94%, the single widest gap in the dataset. Read together, the conversations lean toward making things: writing, building, shipping. The automation-versus-augmentation split is essentially even (49.17% against 50.83%), so these are not conversations that simply hand tasks off; half the time the model is a collaborator, not a replacement.

Coursework and learning conversations make up 23.09% of the sample, against a global 16.45%. That tells us a large slice of usage looks like structured learning — it does not tell us who is typing, and I will not pretend it does. Work-related conversations overall sit at 45.01%, slightly above the global 43.36%, which fits a pattern where learning and building blur into each other rather than sitting in separate boxes.

One honest caveat: this is a single snapshot, not a trend line. It cannot say whether any of these shares are moving. What it can say is that on this snapshot, the Philippines leads Korea in software development, in education and learning (13.02% against 11.21%), and in hobbies and lifestyle conversations (8.85% against 6.44%) — a breadth that spans building for work and building for fun.

That breadth is exactly where I can be useful. If you are in the Philippines and mid-build, I help for free, informally, wherever you are — and the time gap with Seoul is small enough that real-time pairing is easy. My day-to-day work is shipping apps to the App Store and Play Store, wiring up AI agents and MCP tooling, backend and auth systems, mobile development, and the reliability engineering that keeps a thing alive after launch. If your next conversation with a model is a software development one — and the data suggests that is likely — bring the hard part to me instead.

The numbers

PhilippinesGlobal
Anthropic Usage Index0.44 (rank 88 of 121)1.00 baseline
Work conversations45.01%43.36%
Learning / coursework23.09%16.45%
Automation-leaning49.17%48.62%
Augmentation-leaning50.83%51.38%

Most common request topics

TopicPhilippinesGlobalDifference
Content Creation & Copywriting28.82%22.72%+27%
Software Development16.63%11.51%+44%
Education & Learning13.02%13.23%-2%
Hobbies & Lifestyle8.85%9.49%-7%
Research & Intelligence5.8%10.94%-47%

Job-task categories these conversations matched

CategoryShare
Computer and Mathematical26.65%
Arts, Design, Entertainment, Sports, and Media18.99%
Educational Instruction and Library10.88%
Office and Administrative Support7.75%
Sales and Related7.23%

Where Philippines is ahead of Korea

  • Software Development — 16.63% here, 10.42% in Korea (59.6% higher)
  • Hobbies & Lifestyle — 8.85% here, 6.44% in Korea (37.4% higher)
  • Education & Learning — 13.02% here, 11.21% in Korea (16.1% higher)

Ask me something

I’m Seunghan, an engineer in Seoul. I answer questions about shipping apps, AI agent tooling, backends, and reliability — free, and for anyone. Here’s what I can help with and why.

GitHub · KakaoTalk open chat · All countries


Figures from the Anthropic Economic Index, snapshot 2026-05-01. They describe observed Claude conversations matched to job tasks — not who users are, not employment, not the labour market. A country absent from the index means data was not published, never a measured zero.