Does CMU Resolution №875 Actually Fix Defence Tech Export?
TL;DR: CMU Resolution №875, enacted in mid-2026, creates a simplified export permit corridor for Ukrainian defence tech manufacturers — but only for contracts exceeding 15 million UAH, leaving the majority of SME-scale deals in the old, slow lane. The fee structure (0.1–0.5% of contract value) adds a real cost layer that most hardware founders haven’t modelled yet. AI-assisted document compliance tooling is becoming the practical bridge while the regulatory plumbing matures.
At a glance
- Resolution №875 was adopted by the Cabinet of Ministers of Ukraine in 2026 and targets export control simplification specifically for defence and dual-use technology producers.
- 15 million UAH is the minimum contract value to qualify for the expedited permit corridor — approximately €340,000 at the July 2026 NBU reference rate.
- Permit timeline under the simplified regime: ≤10 working days, versus the standard 30+ working days under the previous SSECU procedure.
- Fee range: 0.1% to 0.5% of contract value, with a hard cap of 85,000 UAH per permit application.
- Anastasia Povelko (Juscutum Technology & Investment Practice) identified 3 distinct financial risk zones for manufacturers in her July 15, 2026 column for AIN.ua.
- Ukraine’s State Export Control Service (SSECU) processed over 4,200 export permits in 2025, according to its annual transparency report — a volume that makes manual review a genuine bottleneck.
- The simplified corridor is available to legal entities only — sole proprietors (FOP) are currently excluded.
Q: What does “simplified” actually mean in practice under №875?
The resolution creates a two-track system. Track 1 (standard) stays unchanged — full SSECU review, 30+ working days, legacy documentation pack. Track 2 (simplified) applies when: (a) the exporter is a registered legal entity, (b) the contract value exceeds 15M UAH, and (c) the goods fall under specific commodity codes pre-approved for the corridor.
In concrete terms: a Ukrainian drone component manufacturer shipping a €400K batch to a NATO-country integrator can now submit a streamlined application and expect a decision within two calendar weeks rather than six. That is genuinely meaningful for cash-flow planning — a net-30 buyer relationship becomes viable where it previously wasn’t.
Where we see this hit compliance workflows in production: our docparse MCP server, which we run for several SaaS clients doing cross-border contracting, ingests permit-package PDFs and extracts commodity codes, contract values, and end-user certificates in a single pipeline pass. In June 2026 we measured an average of 2.3 minutes per document for structured extraction versus 45+ minutes of manual paralegal time. The delta matters when you’re preparing a five-document permit package under deadline.
Q: What are the real financial risks manufacturers aren’t seeing yet?
The fee structure is the underappreciated risk. At 0.5% on a 15M UAH contract, you’re looking at 75,000 UAH in permit fees before the goods cross the border — money that’s typically not priced into the original bid because the sales team quoted before legal ran the compliance check.
Povelko’s Juscutum analysis identifies three specific risk zones: (1) fee miscalculation at the contract-structuring stage, (2) contract amendments that push value above or below the threshold mid-process, triggering track switching, and (3) currency-fluctuation exposure between contract signing and permit issuance, since fees are calculated in UAH on the signing-date value.
In March 2026 we ran into a structurally identical problem while building a procurement-compliance workflow for a fintech client dealing with cross-border vendor contracts: the fee wasn’t a percentage of invoice but of the total contract value including options, which doubled the expected cost. We caught it by running contract text through our transform MCP server with a custom extraction schema that surfaces total-commitment clauses separately from base-order value. The same pattern applies directly to defence-tech exporters reading their №875 applications.
Q: Where does AI tooling fit into this compliance picture?
The honest answer is: at the preparation stage, not the decision stage. No AI system is making SSECU permit decisions, and none should. But the document-preparation burden — commodity code classification, end-user certificate formatting, contract-value calculation memos — is exactly the kind of structured, repetitive, high-stakes text work where LLM-assisted pipelines provide asymmetric value.
Our production configuration uses Claude 3.5 Sonnet (via Anthropic API) as the reasoning layer in document-processing workflows. At our measured rate of approximately $0.003 per 1K input tokens for Sonnet, a full permit-package parse — typically 8,000–12,000 tokens of contract and product-spec text — costs under $0.05 per run. That’s not a rounding error when you’re doing 50 permit packages per quarter; it’s a genuine cost model.
The specific MCP servers relevant here are docparse (structured field extraction from PDFs and DOCX), transform (schema-driven reformatting of extracted data into permit-form templates), and knowledge (maintaining a local vector store of SSECU commodity-code definitions and their Ukrainian/English equivalents). Together they form a three-stage pipeline: ingest → classify → format. We have this pattern running in an n8n workflow (internally referenced as the Compliance Doc Prep chain, built on n8n v1.89.2) that reduces per-permit prep time from an estimated 5–6 hours to under 90 minutes for a trained operator.
Deep dive: Ukraine’s export control modernisation in wider context
Resolution №875 doesn’t exist in a vacuum. It’s the downstream output of a multi-year Ukrainian effort to align export control infrastructure with EU standards — specifically the EU Dual-Use Regulation (EU) 2021/821, which the European Commission updated with significant digital-product provisions that directly affect software-enabled weapons systems and AI-driven targeting components.
The EU’s 2021/821 framework, as documented in the Official Journal of the European Union (OJ L 206, 11.6.2021), establishes the concept of “technology transfer” as a controlled export event — meaning that a Ukrainian developer pushing firmware to a foreign hardware partner via GitHub could technically trigger export control obligations depending on the commodity classification. Ukraine’s incorporation of analogous provisions into its own framework is a prerequisite for deeper defence-industrial cooperation with EU member states.
This context matters because it explains why the 15M UAH threshold exists at all. It’s not arbitrary — it mirrors the de minimis thresholds used in several EU member states for simplified procedures under 2021/821, scaled to Ukrainian economic conditions. The Kyiv School of Economics’ defence industry research unit (KSE Institute) noted in its Q1 2026 defence-sector brief that approximately 73% of active Ukrainian defence-tech exporters have average deal sizes below €300K — which sits just under the 15M UAH / ~€340K floor. This means the simplified regime, as currently calibrated, structurally excludes most of its intended beneficiaries.
The broader international comparison is instructive. The U.S. State Department’s Directorate of Defense Trade Controls (DDTC) processes export licence applications under ITAR with a statutory 60-day review target — and routinely exceeds it for complex cases. The UK’s Export Control Joint Unit (ECJU) targets 20 working days for standard individual export licences. Ukraine’s new 10-working-day target for the simplified corridor, if actually achieved at volume, would make it faster than either major Western equivalent. That’s a genuine competitive advantage for Ukrainian defence tech in NATO procurement pipelines — if the SSECU builds the processing capacity to honour it.
The risk is institutional. The SSECU processed 4,200+ permits in 2025 with existing staffing. The simplified corridor, paradoxically, may attract more large-contract applications precisely because it’s faster, increasing total volume while the same headcount absorbs it. Without parallel investment in SSECU digital infrastructure — ideally an API-connected application portal rather than the current PDF-email submission model — the 10-day target becomes a marketing claim rather than an operational reality.
Juscutum’s Povelko correctly identifies the legal risk surface here: if SSECU misses the 10-day window, the resolution’s current text does not provide an automatic deemed-approval mechanism. The applicant’s remedy is administrative appeal, which takes longer than the original review. That asymmetry needs legislative attention.
Key takeaways
- Resolution №875’s 10-working-day permit window is faster than ITAR (60 days) or ECJU (20 days) — if SSECU delivers.
- The 15M UAH threshold excludes ~73% of Ukrainian defence-tech exporters by deal size (KSE Institute, Q1 2026).
- Permit fees of 0.1–0.5% of contract value create a pre-border cost of up to 85,000 UAH per application.
- AI document pipelines (docparse + transform MCP chain) cut permit-prep time from 5–6 hours to under 90 minutes.
- Sole proprietors (FOP) are entirely excluded from №875’s simplified corridor — a structural gap requiring legislative fix.
FAQ
Q: Can a Ukrainian defence-tech startup use №875 if it hasn’t yet made its first export? First-time exporters are not explicitly excluded by the resolution’s text, but SSECU practice has historically required at least one prior registered export transaction as a credibility baseline during review. Legal counsel (Juscutum or equivalent) should be engaged before the first application to document the company’s compliance programme — the absence of an internal compliance officer or written export-control policy has been cited as grounds for extended review in prior SSECU decisions.
Q: Does software count as an export under №875’s scope? Yes, with significant nuance. Software that is “specially designed or modified” for military end-use falls under Ukraine’s controlled-technology list regardless of delivery method — including electronic transmission and cloud access. This directly affects Ukrainian AI/defence-tech companies providing SaaS targeting, logistics optimisation, or battlefield analytics tools to foreign customers. The EU Dual-Use Regulation 2021/821 contains identical provisions, and Ukraine’s framework mirrors that language. Any cross-border software access arrangement for defence clients should be reviewed under №875’s commodity-code definitions before go-live.
About the author
Sergii Muliarchuk — founder of FlipFactory.it.com. Building production AI systems for fintech, e-commerce, and SaaS clients. We run 12+ MCP servers, n8n workflows, and FrontDeskPilot voice agents in production.
Regulatory-compliance automation is one of the highest-ROI applications we’ve deployed for Ukrainian B2B clients navigating cross-border documentation at scale.