About
Built by export control practitioners,
for export compliance teams.
ITAR Screen exists because we spent years on the wrong side of a manual classification process — and built the tool we wished existed.
Why we built this
The problem with manual USML classification
Manual USML classification is slow, expensive, and difficult to audit. The U.S. Munitions List spans 21 categories and hundreds of subcategories — each with specific technical parameters, exclusions, and cross-references to the EAR Commerce Control List. For a company with a product portfolio of any size, producing a documented classification for each item means either retaining specialized export counsel at $300–$600/hour or building an internal team with years of regulatory experience. Most growth-stage companies can't afford either.
The enterprise software market hasn't helped. Legacy platforms like Descartes Visual Compliance and Amber Road (now E2open) were built for Fortune 500 trade compliance departments — multi-year contracts, 6-month implementations, and price tags of $15,000–$100,000/year. The result is a massive gap: well-funded enterprises have tools, and everyone else is screening counterparties in spreadsheets and emailing lawyers for every product classification.
ITAR Screen was built to close that gap. We wanted a tool that returns a USML citation with AI reasoning in seconds, maintains a 10-year immutable screening record that supports 22 CFR §122.5 recordkeeping obligations, and costs less than a single hour of outside counsel. The goal was never to replace export control lawyers — it was to make the work that lawyers review better before it ever reaches them.
The technology
How ITAR Screen works
Claude AI backbone
ITAR Screen is powered by Anthropic's claude-sonnet-4-6 — a frontier AI model evaluated specifically for structured regulatory reasoning. Every classification passes through a purpose-built system prompt that grounds the model against the full USML category text, returning structured output with USML state, category citation, risk level, and reasoned narrative.
45,000+ watchlist entries
Our denied-party screening engine aggregates the OFAC SDN List, BIS Denied Persons List, and Trade.gov Consolidated Screening List — covering more than 45,000 individuals, entities, and vessels. Lists are updated daily via direct government data feeds, not third-party redistributors.
Daily government feeds
USML reference text and all watchlist data are pulled directly from authoritative government sources — DDTC, OFAC, BIS, and Trade.gov. When regulatory text changes, the classification engine is updated and new classifications reflect the current version of the rules.
Versioned classifications
Every classification is timestamped and stored against the specific USML reference version in use at the time of the call. If you classified a product two years ago and the USML category was later amended, your original record remains intact and reflects what the rules said when you ran the screen.
Gideon Dynamics
The company behind ITAR Screen
Gideon Dynamics Corp is a software company focused on regulatory intelligence tools for the defense and dual-use technology sector. We build software for compliance teams that are under-resourced relative to their regulatory exposure — and we measure success by whether those teams can demonstrate a documented compliance posture at audit time.
ITAR Screen is our flagship product. We are based in the United States and our infrastructure runs exclusively on U.S. servers.
Our approach
Designed for human review, not to replace it
When ITAR Screen returns an AMBIGUOUS result, it does not guess. It surfaces the specific uncertainty, identifies which USML categories could apply, and recommends escalating to export control counsel. This is intentional. We believe the right role for AI in export compliance is to handle the clear cases quickly and flag the hard ones clearly — not to paper over genuine ambiguity with a false confidence score.
Every ITAR Screen result includes the AI's full reasoning narrative — not just a result code. This means your legal reviewer can evaluate whether the reasoning is sound, not just accept a black-box output. If the AI made a wrong inference about your product, you will be able to see exactly where and correct the record before it becomes a compliance problem.
Disclaimer: ITAR Screen provides AI-assisted screening only. Results are not legal determinations and do not constitute export control legal advice. Consult qualified export control counsel before making any export compliance decisions, commodity jurisdiction requests, or regulatory filings with DDTC or BIS.