OFAC Screening Benchmark

A benchmark is a standardized test of a screening tool's accuracy, speed, and false positive rate against known SDN entries.

TL;DR

TL;DR: An OFAC screening benchmark is a standardized test of a screening tool's accuracy, speed, and false positive rate, run against a known set of SDN entries.

What a benchmark measures

A useful benchmark has three parts. Accuracy measures whether real SDN matches are caught. Speed measures how long a check takes. The false positive rate measures how many clean names are wrongly flagged. All three matter because a tool that is fast but misses matches, or accurate but slow, fails in production where a payment waits on the answer.

Why it matters for agent payments

An agent pays in real time, so speed and accuracy must hold together. SanctionsAI is designed for that balance: a single HTTP call returns ALLOW or BLOCK in under 100 ms against 19,218 SDN names and 947 OFAC-listed crypto wallets, synced hourly so the benchmark reflects live data, not a frozen sample.

How to run your own

Take a known set of SDN entries plus a set of clean names, then measure hits, misses, and false flags. Because OFAC applies strict liability and penalties start at $356,000 per violation, the cost of a missed match is far higher than the cost of a slow benchmark run. A benchmark also gives you a baseline to catch regressions, so a future change that hurts accuracy shows up before it reaches production. Measuring once is a test; measuring repeatedly is a safeguard.