Kyrgyzstan has been running national digital skills programmes for almost a decade. They have names that change with each administration — Taza Koom, Sanarip Kyrgyzstan, Digital Kyrgyzstan — but the underlying ambition has been continuous: build the digital literacy of the population, train teachers, equip schools, reach civil servants, push training out into rural communities. Across the same window, the country has not built one thing that ought to come with that ambition. There is no national system that measures, in any standardised way, whether the digital skills of the population are actually rising.
Neither the Ministry of Education and Sciences nor the National Statistics Committee conducts regular tracking of digital literacy levels. What exists is a patchwork of donor-funded surveys (UNDP, World Bank, ITU, each using different methodologies), academic snapshots of specific populations, and ad-hoc project evaluations. None of them produces a comparable, longitudinal, country-wide read on whether population-level digital skills are improving or by how much. The programmes scale; the yardstick that would tell us whether they are working has never been built.
The work that did happen — at the layer below measurement
To say the country has no measurement system is not to say nothing has been built. The training side of the picture has, in fact, scaled steadily and across multiple modalities at once.
- Learning for Future (World Bank-funded) has trained 40,000 teachers in digital tools and distributed 24,000 computers across roughly 50% of public secondary schools. The Ministry of Education is working to extend the device coverage and is in the process of assembling around $65 million in additional grants and loans for K–12 reforms.
- Sanarip Insan (EU-funded) — the project I orchestrated through the Internet Society Kyrgyz Chapter (ISOC KG). It produced the Kyrgyz/Russian translation and local adaptation of Microsoft’s Digital Literacy Framework and the GSMA Mobile Internet Skills Training Toolkit (MISTT), distributed the framework across all 2,300 secondary public schools, integrated it into the curricula of 12 universities, and disseminated MISTT through civic education centres, libraries, and local administrations — enriched with locally relevant modules including ones built around Tunduk.
- The University of Central Asia, operating as a Digital CASA subcontractor, has trained 1,500 civil servants in digital tools.
- Grassroots organisations — Leader NGO, ISOC KG, Enactus Kyrgyzstan — have run practical digital literacy training in rural areas, in the Kyrgyz language, filling the parts of the country that formal institutions are slowest to reach.
Sanarip Insan, from inside
The part of this picture I have direct, hands-on experience with is Sanarip Insan («digital citizen» in Kyrgyz). I orchestrated its implementation through ISOC KG — the translation work, the local adaptation, the dissemination architecture across schools, universities, libraries, civic education centres, and local administrations. The choice to anchor the project in two globally tested instruments, Microsoft’s Digital Literacy Framework and the GSMA MISTT, was deliberate: both are validated curricula that allow Kyrgyz training output to be benchmarked against international peers, rather than designed in isolation.
The substance of the localisation work — what to keep, what to adapt, where to add Kyrgyz-specific modules around government services like Tunduk, how to structure the cascade from civic centre trainer to end learner — drew on prior research on digital skills and competencies that I had built up over the years before the project, and on operational ICT experience from running KG-IX and adjacent infrastructure work. It is one of the cases where research and infrastructure work compound rather than running on parallel tracks.
What the rollout illustrated, in microcosm, is the same gap that runs through the country’s wider digital skills picture. It is possible to translate a globally tested toolkit into Kyrgyz and Russian, ship it through 2,300 schools, embed it into 12 universities, and disseminate it through every layer of civic infrastructure that will accept it. It is much harder — without a national measurement framework — to say with any confidence whether the people who went through the training are doing things differently three months, six months, or two years later. The training is the input that programmes report on. The behavioural change is the output that, at the population scale, no one is measuring.
The school computer count, and what’s actually in the lab
One of the headline indicators reported for school-level digital readiness is the count of computers in secondary schools. On paper, most secondary schools across the country report some level of computer equipment. The official totals look respectable. The picture in the rooms where those computers actually sit is not the same picture.
I have personally visited many secondary schools across Kyrgyzstan — in Batken Oblast in the south, Issyk-Kul Oblast in the east, and Chui Oblast in the north. The geographic spread is wide; the picture is the same in each of them. A material share of the computers being counted as school equipment are fifteen years old or older. They are still on the institutional balance sheet because the formal liquidation process for state-registered equipment is hard and slow — not because they are running anything that resembles modern educational software. They cannot run current operating systems, current browsers, or any of the curricula that the digital skills programmes assume in the room.
The result is a measurement problem layered on top of a measurement problem. The country does not have a population-level digital literacy index. It also does not have a clean count of how much of its school computer stock is actually capable of supporting the digital training the same schools are reporting on. The phantom-equipment effect is invisible at the policy dashboard level, immediate the moment a teacher tries to open a modern application in a Year 7 classroom.
The teacher paradox is a measurement problem too
One of the better-documented findings in the DECA field research is that Kyrgyz teachers self-report a high level of basic digital literacy, and that this self-assessment does not translate into classroom practice. Teachers are equipped with laptops; adoption and active use of those laptops is low. Educators understand basic tools but lack pedagogical frameworks for integrating them into a lesson plan. School connections to high-speed internet are inconsistent even where the signal exists in the district. And — most awkwardly — teachers are prohibited from using smartphones in classrooms, which means the device most teachers and students are most fluent in is officially the one device that cannot be part of the lesson.
The teacher paradox is usually framed as a pedagogy or policy problem. It is also, more quietly, a measurement problem. Without a way of tracking which teachers actually use which tools in which lessons over time, a self-reported literacy score remains the only national signal — and it does not survive contact with the classroom.
One thing that does get measured: the .kg domain
In the absence of a population-level digital literacy index, indirect indicators take on more weight than they should. One of the most telling, because it does get measured, is the count of registered .kg domains. Over the three years to 2024, that count dropped from 8,373 to 5,040. Only about 25% of content on .kg sites is in the Kyrgyz language. The Kyrgyz Wikipedia, the most-cited proxy for online language vitality, holds roughly 100,000 articles against millions in Russian and Uzbek.
None of those numbers, on their own, prove a digital literacy story. Together, and read against the country’s training output, they suggest a working hypothesis: the supply of training is real, but the digital environment people are being trained into has been thinning rather than thickening — fewer Kyrgyz-language sites to read, fewer .kg-hosted services to interact with, and an internet whose richest content sits in languages that not all of the trained population can fluently use.
Where Central Asia compares
| Country | UNDP Digital Literacy & Skills Score (out of 5) |
|---|---|
| Kazakhstan | 3.87 |
| Kyrgyzstan | 3.44 |
| Uzbekistan | 3.23 |
| Tajikistan | 3.07 |
Within Central Asia, Kyrgyzstan sits in the middle of the band — ahead of Tajikistan and Uzbekistan, behind Kazakhstan. The relative position is consistent with what the field shows: meaningful programme activity, real institutional commitment across multiple administrations, but no clear way to demonstrate progress over time. Kazakhstan’s value, in this comparison, is less the score itself and more the fact that it does collect standardised ICT indicators disaggregated by region and by population group, and reports them publicly. Kyrgyzstan has the institutional capacity to do the same. It has not yet been done.
The advanced-skills floor
For the population already past basic literacy and looking for industry-level competencies — data analysis, cybersecurity, AI/ML, digital entrepreneurship — the system thins out quickly. Roughly 14% of those who seek advanced digital skills training can access a high-level programme. Most of what does exist is concentrated in Bishkek and runs through a small number of private or internationally supported providers. University curricula deliver basic computer literacy reliably; industry-focused content is uneven.
The pipeline question that comes out of this is also a measurement one. Female ICT graduates outnumber male graduates in some institutions, but women still make up only around 28% of the ICT workforce. The training output and the workforce composition are saying two different things, and the country does not have the longitudinal employment-by-graduate-cohort data that would let anyone resolve which one is the closer description of reality.
What is built, and what is not yet built
Read across all of this, the digital skills picture in Kyrgyzstan is more durable than the headline rankings suggest. Successive administrations have kept the work going. Donor co-investment has been steady. Grassroots organisations have filled the parts of the country slower-moving institutions reach. Toolkits have been localised into the languages the population actually speaks.
The programmes have scaled. The instrument that would tell us whether they have shifted the population — has not.
The next pieces in this series move from the supply side of the digital landscape — infrastructure, exchange, skills — toward who is left behind even where signal, devices, and training are present. The digital gap, properly understood, sits at the intersection of all three.
The Digital Skills section of the assessment for the Kyrgyz Republic was prepared by Aziz Soltobaev as part of the DECA assessment. The view in this piece draws on direct involvement in the orchestration of the EU-funded Sanarip Insan project through the Internet Society Kyrgyz Chapter (ISOC KG) — including the Kyrgyz/Russian translation and adaptation of Microsoft’s Digital Literacy Framework and the GSMA Mobile Internet Skills Training Toolkit — and on direct school visits across multiple oblasts. The work builds on prior KG Labs research on digital skills and competencies and on operational ICT experience from KG-IX and adjacent infrastructure work. For an updated read on Kyrgyz digital literacy programmes, school-equipment baselines, or a tailored briefing on the .kg domain and Kyrgyz-language content layer, get in touch.