Ukraine Launches Diia.AI: A National AI Assistant for Government Services
Ukraine has rolled out Diia.AI, the first national AI-based assistant for public services. The goal is clear: move from a digital state to an "Agentic State," where AI doesn't just answer questions but takes action on behalf of citizens.
First Deputy Prime Minister - Minister of Digital Transformation Mykhailo Fedorov outlined the direction: practical AI at scale, with an ambition to rank among the top three countries in applied AI. This is a move from pilots and chatbots to production-grade automation.
What Diia.AI Does Right Now
More than 85% of citizens' requests for core services in the Diia app are processed by AI assistants. Launched in September 2025, the project was recognized in the World Records Book under "Firsts" and "Inventions," as shared by project CEO Anna Krysiuk during an international conference.
The result: faster turnaround, fewer queues, and consistent service quality at national scale. It also sends a signal to investors that Ukraine is building high-value, AI-enabled infrastructure.
From Digital State to Agentic State
Ukraine's next step is shifting from self-service portals to AI agents that complete tasks end-to-end. Think filing, checking status, updating records, and nudging the user only when needed.
This is a policy and architecture shift: services are designed around intents (birth registration, benefits, permits), not forms and pages.
Partnerships Powering the Build
Ukraine is partnering with global tech communities and governments, including collaboration with Google to develop government AI products. These alliances accelerate delivery and reduce technical risk.
For context on the ecosystem, see the Ministry of Digital Transformation of Ukraine official site and Google's AI capabilities for public-sector workloads on Google Cloud.
Why This Matters for Government Leaders
Most public-sector AI efforts stop at chatbots. Ukraine moved to agents that execute. The takeaway: citizens want outcomes, not conversations.
If you lead a department, this is a clear blueprint for switching from "digital forms" to "AI-completed services."
A Practical Playbook to Adapt
- Start with the top 10 high-volume, rules-based services (licenses, benefits, certificates). Define the user intent and success criteria for each.
- Deploy AI as the first line for triage, data prefill, and status updates. Route edge cases to humans with full context.
- Adopt agentic workflows: the AI initiates, tracks, and completes steps across systems via secure APIs.
- Use staged rollouts: sandbox → limited pilot → nationwide scale. Measure completion time, error rates, and satisfaction at each stage.
- Stand up a cross-functional "AI Services Team" (policy, legal, IT, service owners, security) with weekly ship cycles.
Governance and Risk Controls
- Data minimization by default. Only the attributes required for a decision are processed.
- Human-on-the-loop for sensitive cases. Every automated decision is auditable with clear logs.
- Model policies: versioning, retraining cadence, prompt/response retention rules, and red-teaming.
- Security: zero-trust access, encrypted data in transit/at rest, and per-service API scopes.
- Fairness: test for disparate impact on language, region, and demographics; publish summaries.
KPIs Worth Tracking
- AI-handled rate by service (target >70% for mature flows).
- Median time-to-completion vs. legacy channels.
- First-contact resolution and re-open rates.
- Cost per completed case and projected savings.
- Appeals/complaints per 10,000 decisions.
Procurement and Partnerships
- Mix commercial AI platforms with sovereign control: clear data residency, model isolation, and exit options.
- Mandate open standards for identity, payments, and document exchange to avoid vendor lock-in.
- Negotiate outcome-based contracts tied to service-level improvements, not just licenses.
What's Next
Ukraine plans to scale agentic services across more life events and deepen partnerships to keep pace with demand. The focus is practical implementation and measurable outcomes.
If your agency is mapping a similar path, align on one principle: build for outcomes, instrument everything, and let AI handle the busywork so staff can focus on judgment and oversight.
Further Resources
- Ministry of Digital Transformation of Ukraine: thedigital.gov.ua
- Public-sector AI training and role-based learning paths: Complete AI Training - Courses by Job
Your membership also unlocks:
