Nepal's AI moment in 2026: the policy, the money, and what it means for builders

Something shifted in Nepal's tech story this year. For a long stretch, "AI in Nepal" meant a panel at a conference and a handful of outsourcing contracts. In 2026 it means a national policy on the books, more than $200 million moving into local companies, and 25 founders on a plane to Chennai. The pieces are finally in the same room.

We build here, so we read these signals closely — not as headlines, but as a to-do list. Here's what actually changed, and what we think it means if you're shipping software from Nepal.

The policy is real. The roadmap is still thin.

In August 2025 the government approved the National AI Policy — its first. It sets up an AI Regulation Council chaired by the communications minister, a National AI Centre to coordinate development, and a target of training 5,000 AI professionals within five years. It points AI at the sectors that matter for a country like ours: health, education, agriculture, and public administration. There's even a two-year review cycle baked in, which is a quietly sensible thing for a field that moves this fast.

It's a genuine step. It's also fair to say the policy is stronger on intent than on execution — analysts have flagged the absence of a concrete implementation roadmap, thin AI infrastructure, and data-governance rules that predate the problem they now have to solve. A policy is a direction, not a delivery. The interesting question is who turns it into working systems.

The money showed up too

Policy alone doesn't build companies; capital does. The Silicon Peaks ecosystem now counts 94+ companies — SecurityPal, Fusemachines, CloudFactory, Baato and others — with over $200 million deployed into them, from early venture to development finance. In the background, Nepal's IT exports crossed the billion-dollar mark for the first time in 2025. That's not a rounding error; it's an industry.

The pipeline is filling, too. The second India–Nepal Startup Partnership Network cohort sent 25 startups — chosen from 380 applications — to IIT Madras for an eight-week programme this June. From the first cohort of 24, nine walked away with incubation or investment offers. And on the government's own side, a domain-aware AI platform called SITA is being used to turn national datasets into briefs in a fraction of the usual time.

What this means if you're building here

Strip away the announcements and a short, practical list remains:

• Ship applied AI now — the policy direction is set and won't reverse. The advantage goes to teams with something running in production, not a pitch deck.
• Build evals and guardrails in from day one. A regulation council means accountability is coming; treat it as a feature you already have, not a cost you'll bolt on.
• Apply to the ecosystem programmes. IN-SPAN seats are scarce — 25 from 380 — but the incubation and investment offers behind them are real.
• Make data governance a selling point. The national frameworks are dated, so clean data handling is a way to stand out, not just a box to tick.

The real risk for Nepal isn't moving too fast on AI. It's the policy staying on paper while the talent boards a flight and doesn't come back. The antidote is unglamorous: teams that ship real things, here, for clients who pay. That's the whole game — and it's the work we get up for.