Top 74%
in this catalogue (1041 scored practices)
Scores cluster high, so position within the catalogue is often more telling than the number alone.
Evidence of impact / measured public value2/3
Two independent outlets corroborate concrete usage figures (150,000+ uses in four months, 2 million-plus Q1 requests, 30% drop in in-person visits) from named officials, though the figures are self-reported by government and not independently audited.
Transparency, fairness & accountability1/3
Officials publicly disclosed scope (50 services, ~30% of requests) and an explicit advisory-only boundary for sensitive decisions, but no accuracy, error-rate, or bias audit has been published.
Transferability / demonstrated replication1/3
The assistant runs on Kazakhstan's own sovereign infrastructure (national supercomputer, AlemLLM), so replicating it elsewhere requires comparable domestic infrastructure investment, not just the concept.
Scalability beyond pilot2/3
Government has a stated, funded roadmap to expand from 50 to roughly 200 services and from 30% to 70% of citizen-request coverage within the same year, backed by already-large usage volumes.
Governance, capability & sustainability1/3
A deliberate design choice keeps the AI advisory-only for benefit eligibility, with humans retaining final decisions, but there is no published independent oversight, appeals mechanism, or audit of the system.
Evidoria. eGov GPT — Kazakhstan's Conversational AI Assistant Cuts In-Person Service-Centre Visits by 30%. Persistent ID: e411324c-4a81-4b80-b57e-c5328ea1b9cd.