How AI Is Helping Government Companies in Ukraine Cut Costs and Improve Efficiency

AI is already reshaping Ukrainian government and public services: Ukraine WINWIN 2030 AI strategy treats AI as a strategic sector and scales GovTech (Diia serves over 22 million users), while Kyiv's pragmatic push to adopt commercial AI and regulatory sandboxes speeds automation, cuts routine labor, and compresses analysis cycles.

Multisensor platforms - Delta, Griselda and Zvook - illustrate big efficiency gains (Griselda can reduce human labor by up to 99%), and observers at CSIS analysis of Ukraine's military AI ecosystem note government focus on rapid, mission-driven deployment.

The result: lower costs and faster services for citizens, if energy limits, ethical safeguards, and workforce reskilling keep pace.

Bootcamp	Length	Early-bird	Register
AI Essentials for Work	15 Weeks	$3,582	Register for AI Essentials for Work - 15-week bootcamp

On the battlefield I did not see a single Ukrainian soldier. Only drones. I saw them [Ukrainian soldiers] only when I surrendered. Only drones, and there are lots and lots of them. Guys, don't come. It's a drone war.

Across Ukraine the public sector and defense have moved from pilots to purpose-built AI tools that shave time and cost from both citizen services and combat operations: e‑government platforms like Diia digital government platform (serving over 22 million users) expanded by using startup-style MVPs and rapid iteration, while a growing defense ecosystem - supported by state-backed accelerators such as Brave1 - has marshalled hundreds of startups and volunteer projects into operational systems that fuse drone, satellite and open‑source data.

Independent analysts find most military AI remains “supportive” rather than fully autonomous, yet practical gains are striking: situational platforms compress decision cycles from hours to under 30 minutes (tactically as low as five), and intelligence pipelines like Griselda can turn raw feeds into actionable alerts in about 28 seconds.

These shifts reflect complementary strengths - public legitimacy and scale plus private‑sector speed - but also expose gaps in compute, long‑term strategy and regulatory safeguards that Ukraine is now addressing through ethical initiatives and soft, sandboxed governance designed to keep innovation accountable and interoperable with allies (CSIS analysis of Ukraine military AI ecosystem, Friends of Europe analysis of startup-driven defence innovation in Ukraine).

Indicator	Snapshot
Diia users	~22,000,000 (WEF)
AI companies in Ukraine	243 (industry reports)
Brave1 grants	299 grants totaling ≈ USD 6.5M (CSIS)
Drone procurement (FY 2024)	UAH 58.8 billion (CSIS)

“Artificial intelligence standards are a matter of security for millions of Ukrainians. Soon AI will be in Diya and Dreams, so it is important to define the principles of safe use of this technology in our products.”

Diia.AI is turning Ukraine's superapp into an “agentive” state helper: rolled out in open beta as the world's first government AI that actually completes requests, it lets users get an income certificate and other services through a simple chat - no long forms, just plain language - and works 24/7 to pull registry data, suggest the right programs, and guide next steps; testers can already type a life-situation prompt like “my house was damaged by shelling” and receive a tailored recommendation (eRestoration) and how to apply, a vivid leap from menus to meaningful automation.

Built on Diia's existing digital infrastructure (millions of users and legal digital IDs) and supported by programs such as UK DIGIT and Swiss‑Ukrainian EGAP, the assistant is intended to scale into social, veterans', notary and business services while the Ministry explores monetisation and a spin‑off model.

Read the official launch details and the Diia overview for how that operational scale makes 24/7 automated service delivery possible: see the Diia.AI launch report and the Diia ecosystem overview.

Metric	Value / Note
Diia ecosystem users	more than 14 million (Diia)
Digital documents in app	11 digital documents (Diia)
Web services	70+ online services (Diia portal)
Diia.AI initial live service	Income certificate via chat (beta)

Starting today, the first service – obtaining an income certificate – will be available, said Digital Minister Mykhailo Fedorov.

WINWIN 2030 turns Ukraine's wartime innovation impulse into a blueprint for long-term efficiency: the strategy elevates GovTech and AI into national priorities, stitches together 14 strategic sectors (from DefenseTech to MedTech and Semiconductors), and builds institutional muscle through Centers of Excellence so ideas move from pilot to public service at scale - the first WINWIN CoE for Artificial Intelligence is already operational and two more are set to launch by year‑end, creating hubs that combine government, academia and industry.

WINWIN's delivery-first approach - documented in the Ministry's overview and analysis - pairs regulatory simplification, pilot projects and talent pipelines so GovTech can be exported as soft power and Diia-style services can be replicated abroad; the strategy also explicitly frames the state as a market architect that unlocks private initiatives and fast tracks safe testing and market access.

For practical details and the government's framing of WINWIN as a platform for global partnership, see the official WINWIN site and the Ministry's analysis of the strategy.

Element	Note
Centers of Excellence	AI CoE operational; two additional CoEs launching by year‑end (WINWIN)
Priority sectors	14 sectors including GovTech, AI, DefenseTech, MedTech, Semiconductors, GreenTech
GovTech goals	Create world's most advanced digital governance system; export GovTech solutions and host a Global GovTech Centre/Board
Key partners	Ministry of Digital Transformation, Ministry of Education and Science, international partners (e.g., USAID, UK programs)

Ukraine's secret to shaving months - or even years - off fielding cycles is a practical stack of platforms, sandboxes and accelerators that stitch frontline needs directly to makers: Brave1's grant and scouting engine (more than $30M across 500+ projects) fast‑tracks R&D and even compresses NATO Stock Number approvals to months - or a record nine days - while the Brave1 Market acts like an “Amazon for drones and EW,” letting units compare products and get a product card up in hours for rapid purchases; read more on Brave1's program and speeded NSN process Brave1 profile and program overview and on how procurement timelines were rewritten in practice in the CSIS analysis of Ukraine's acquisition reforms CSIS analysis of Ukraine's military acquisition system reform.

Complementary digital tools - DOT‑Chain's marketplace pilots, testing ranges like Iron Range/Zaliznyy Polygon, and reporting apps such as Army+ - close the loop so an urgent front‑line problem becomes a funded commercial solution in weeks; see how the Brave1 Market connects soldiers and makers in real time how the Brave1 Market works: “Amazon for drones and EW” analysis.

The upshot is vivid: what used to be a multi‑year procurement treadmill can now deliver a battlefield‑proven drone or EW kit in the time it takes to train a new operator, cutting cost, risk and bureaucratic drag.

Platform	Function / Impact
Brave1	R&D grants (> $30M, 500+ projects), fast NSN approvals (record 9 days)
Brave1 Market	Online procurement marketplace (≈3,600 products), product cards created in hours
DOT‑Chain	Defense procurement marketplace pilot enabling unit-level orders (pilot funding allocations)
Iron Range / Zaliznyy Polygon	Government testing ranges with digital applications for rapid trials

BraveTech EU is a new strategic step for development of EU's and Ukraine's defence industry cooperation.

Ukraine's battlefield efficiency now pivots on a tight loop between multisensor fusion, rapid text-and-audio mining, and acoustic detection: the Delta platform stitches video, satellite, drone and human reports into a live common operating picture, Griselda slurps unstructured chats and intercepted audio to surface leads that cut analyst time dramatically, and Zvook's acoustic nodes “hear” threats across thousands of square kilometers and hand detections to Delta in seconds - acoustic ranges reach up to ~4.8 km for drones and ~6.9 km for cruise missiles with ~12 seconds to Delta and false positives near 1.6% (CSIS analysis of detection performance).

Together these tools shrink the time from raw feed to action, let computer vision like the Avengers classifier flag vehicles in a couple of seconds, and automate workflows that CSIS finds can reduce human labor in key intelligence tasks by as much as 99%; the practical payoff is simple and stark: commanders get faster, higher‑confidence alerts and analysts focus on verification and judgment rather than sifting noise.

For how these systems fit into Ukraine's military AI architecture and the evidence behind their performance, see the CSIS military AI ecosystem overview and the CSIS analysis of Ukraine's autonomous-warfare capabilities.

System	Key metric / role
Delta	Multisensor common operating picture (video, sat, acoustic, text)
Griselda	Automates unstructured data analysis - up to ~99% labor reduction (CSIS findings on automation benefits)
Zvook	Coverage ≈20,000 sq km; detection: drones ~4.8 km, cruise missiles ~6.9 km; ~12s feed to Delta (CSIS report on Zvook acoustic detection)
Avengers (CV)	Vehicle recognition ~70% in ~2.2s (Ukraine Ministry of Defence report / Diia Digital State platform)

On the battlefield I did not see a single Ukrainian soldier. Only drones. I saw them [Ukrainian soldiers] only when I surrendered. Only drones, and there are lots and lots of them. Guys, don't come. It's a drone war.

Ukraine's push toward low‑cost, on‑device autonomy is turning everyday FPV and tactical drones into resilient, battle‑proven tools: a Pentagon contract will deliver 33,000 of Auterion's Skynode/Skynode strike kits this year - massively scaling what had been incremental autonomy deployments - and domestic kits like ZIR and the VGI‑9 optical guidance module add compact, affordable ATR and last‑mile navigation that resist jamming and let crews train in hours.

These small, hardware‑agnostic AI modules run on inexpensive chips, enable target recognition and tracking out to roughly 1–2 km in optimal conditions, and plug into the Delta/ISR picture so human commanders retain oversight while routine sensing, tracking and final approach are handled on‑device; the result is fewer repeat sorties, higher first‑shot success, and lower logistics burden because software and local manufacture replace long supply chains.

For quick reads on the scale and technical features, see Auterion's Skynode announcement and CSIS's analysis of Ukraine's autonomous‑warfare trajectory, and the VGI‑9 overview for optical guidance specifics.

System	Key metric / role
Auterion Skynode	33,000 strike kits contracted (Pentagon / Auterion)
ZIR autonomy kit	Onboard detection ≈1 km; autonomous engagement up to ≈3 km (CSIS)
VGI‑9	Optical guidance; effective visual range 100–2,000 m; cruise control, EW resilience (VGI)
ATR / On‑device AI	Targeting up to ~2 km; enables last‑mile navigation and reduced sortie rates (CSIS)
2024 procurement	~10,000 AI‑enhanced drones purchased (CSIS)

“We have previously shipped thousands of our AI strike systems to Ukraine, but this new deployment increases our support more than tenfold. This is not only significant for Ukraine's defense but also represents a substantial advancement in drone warfare technology.”

Domestic scaling is the lever that turns battlefield ingenuity into cheaper, faster tools: Ukraine's sprawling network of startups, volunteer workshops and small manufacturers has driven production from bench prototypes to mass output - roughly 2 million drones in 2024 - with CSIS noting about 96.2% of UAVs used that year were domestically produced, so local factories and modular AI kits shave cost by replacing long supply chains and reducing logistics overhead.

The momentum has attracted big money: the EU moved to frontload roughly €6 billion (≈$7 billion) to supercharge mass production, and private investors have poured growth capital into dozens of firms, helping turn iterative, soldier‑driven designs into assembly‑line reality.

The result is lower per‑unit prices, faster replenishment cycles and software‑first upgrades that cut lifecycle costs - one vivid measure: a small workshop's overnight tweak can save the state the time and expense of a multi‑month procurement.

Metric	Value / Source
EU drone funding	€6 billion (~$7 billion) to boost mass production (Business Insider report on EU €6B funding to boost Ukraine drone production)
2024 drone production	~2,000,000 units (ComplexDiscovery analysis of Ukraine's 2024 drone production)
Domestic production share (2024)	~96.2% domestically produced (CSIS analysis on Ukraine's AI-enabled autonomous warfare and domestic UAV production)
Private investment (2024)	>$20M into defense tech startups (ComplexDiscovery)
2024 AI-enhanced drone procurement	~10,000 purchased (CSIS)

"Ukraine has the ingenuity. What it needs now is scale." - European Commission President Ursula von der Leyen

Measured gains from AI in Ukraine are already concrete: intelligence pipelines like Griselda can cut analyst labor by up to 99%, multisensor platforms such as Delta compress decision cycles from hours to under 30 minutes (tactically as low as five), and domestic scale - roughly 2 million drones produced in 2024 with ~96.2% made locally - drives steep per‑unit cost declines and faster replenishment; these findings are documented in CSIS's technical analysis of Ukraine's AI-enabled warfare and industrial scaling.

Civilian and governance examples show similar payoffs: rapid cloud migration and modernization at PrivatBank (migrated to AWS in less than 43 days and improved total cost of ownership within 90 days) kept critical services online while cutting operational overhead.

Complementary research from CEPS highlights Ukraine's strong AI publication and startup base, underscoring how local talent and research density turn battlefield and public‑service pilots into scalable savings and exportable GovTech.

The practical “so what” is striking: a frontline workshop's overnight tweak can replace a multi‑month procurement cycle, saving time and taxpayer money while accelerating capability delivery - proof that measured efficiency often arrives where modular AI meets local production and fast procurement.

Metric	Value / Source
Griselda labor reduction	Up to ~99% (CSIS)
Decision‑cycle compression (Delta)	From hours to <30 min; tactically ≈5 min (CSIS)
2024 drone production	~2,000,000 units; ~96.2% domestic (CSIS / industry)
PrivatBank cloud migration	Migrated to AWS in <43 days; improved TCO in first 90 days (AWS case study)

On the battlefield I did not see a single Ukrainian soldier. Only drones. I saw them [Ukrainian soldiers] only when I surrendered. Only drones, and there are lots and lots of them. Guys, don't come. It's a drone war.

Despite clear efficiency gains, Ukraine's AI rollout faces hard limits and ethical trade‑offs that could blunt long‑term benefits: analysts warn of scarce computing infrastructure and a shortage of skilled staff (many displaced or working remotely), leaving research centres and defense labs starved for GPUs and secure environments (CSIS analysis of Ukraine's military AI ecosystem); education and institutional readiness also lag, with BDO identifying hurdles in schools and universities that must be closed to sustain the talent pipeline (BDO Ukraine report on digital shift in education and AI readiness).

Fragmented strategy, permissive wartime procurement, and vast reconstruction demands - hundreds of thousands of damaged facilities and urgent municipal finance gaps - raise governance, data‑sharing and export‑control risks that could entrench short‑term fixes over durable, accountable systems (Kinstellar analysis of rebuilding Ukraine's post‑war infrastructure).

Mitigation means investing compute, formalising human‑in‑the‑loop rules, and targeting funds so AI scales safely rather than simply faster.

Constraint / Risk	Evidence / Source
Insufficient compute & skilled staff	CSIS: limited computing infrastructure; workforce displacement
Education & readiness gaps	BDO: schools/universities face digital shift hurdles
Reconstruction & funding pressure	Kinstellar: destroyed infrastructure; urgent financing needs
Strategic fragmentation & ethical risk	CSIS: no unified long‑term military AI strategy; permissive regulation

“On the battlefield I did not see a single Ukrainian soldier. Only drones. I saw them [Ukrainian soldiers] only when I surrendered. Only drones, and there are lots and lots of them. Guys, don't come. It's a drone war.”

Policy choices now should lock early gains into durable institutions: fast‑track the Science City legal framework to turn university labs into innovation hubs with clear incentives (including 0% VAT and 0 UAH customs duty for eligible research equipment) and train technology‑transfer managers to move R&D from bench to market (Science City legal framework details); align WINWIN 2030 delivery mechanisms with EU and donor programs so Centers of Excellence scale pilot projects into public services and exportable GovTech; coordinate international assistance through the newly formed International Coalition so reconstruction grants and technical assistance target the estimated €1.2 billion R&D rebuild needs and seed an applied research agency and open data repositories (International Coalition to support Ukraine research and innovation).

Fiscal and regulatory levers matter: keep promoting Diia.City/industrial‑park incentives to attract VC and manufacturing, pair EIC and EU seed funds with matched national grants (the EIC's €20M window shows how targeted capital accelerates scaling), and formalize human‑in‑the‑loop rules, secure compute access, and university financial autonomy so talent stays and innovations mature into low‑cost, government‑grade systems (Diia.City and R&D incentives for Ukraine).

The payoff is practical: with these steps, overnight campus prototypes can be turned into mass‑produced public goods instead of one‑off demos, shrinking time‑to‑benefit for citizens and the front line.

Priority	Target / Example
R&D rebuild funding need	~€1.2 billion (restoration + researcher support)
EIC support	€20 million for Ukrainian deep‑tech SMEs
Science City incentives	0% VAT; 0 UAH customs duty for eligible equipment

“Ukrainian scientists, engineers, and entrepreneurs are showing exceptional resilience. Many continue their work under Russia's daily attacks or displacement. The European Commission does not only express its admiration - we offer them full support – because research and innovation must be part of Ukraine's recovery.”

AI's practical payoff for Ukraine is clear: modular autonomy, multisensor fusion and massive domestic production have turned pilots into enduring efficiencies that strengthen both services and sovereignty - CSIS technical analysis on battlefield autonomy and decision cycles documents how tools like Griselda and Delta shrink analyst workloads (Griselda can cut labor by ~99%) and compress decision cycles from hours to under 30 minutes, while domestic manufacturing produced roughly 2 million drones in 2024 to sustain rapid replenishment and lower unit costs.

Yet gains depend on choices: human‑in‑the‑loop rules, secure compute, ethical oversight, and talent pipelines that keep capability onshore rather than outsourced to risky suppliers.

That's why reconstruction and public‑service use cases - demining, claims processing and Diia‑style automation - must pair fast pilots with transparent audits and workforce reskilling so savings reach citizens, not just balance sheets; practical training like the AI Essentials for Work bootcamp registration can help public servants and entrepreneurs apply prompts and low‑cost AI tools responsibly.

For a technical read on battlefield autonomy and the policy tradeoffs, see the CSIS study on battlefield autonomy and for reconstruction and public‑service applications consult HROMADA's reconstruction and planning overview.

Metric	Value / Note
Griselda labor reduction	Up to ~99% (CSIS)
2024 drone production	~2,000,000 units (CSIS)
ATR engagement range	Up to ~2 km in optimal conditions (CSIS)
Decision‑cycle compression (Delta)	From hours to <30 min; tactically ≈5 min (CSIS)

On the battlefield I did not see a single Ukrainian soldier. Only drones. I saw them [Ukrainian soldiers] only when I surrendered. Only drones, and there are lots and lots of them. Guys, don't come. It's a drone war.