AKRUM logo
Crypto Regulatory

AI Crypto Compliance Agent

Jurisdictional compliance for on-chain transactions and smart contracts. FATF Travel Rule, sanctions screening, and contract risk in one API.

Overview

What it does

Evaluates blockchain transactions, wallet addresses, and smart contract code against the regulatory frameworks of every jurisdiction you operate in. Returns a structured verdict.

Combines on-chain analytics (counterparty risk, mixer interaction, sanctions exposure) with off-chain regulatory context (FATF Travel Rule thresholds, MiCA categorization, US OFAC screening, jurisdictional licensing requirements).

Includes a smart contract auditing pass that flags known vulnerable patterns (reentrancy, integer overflow, unbounded loops, missing access control) and ties findings to specific regulatory exposure.

How it works

A 4-step evaluation flow

01

Submit payload

POST transaction OR wallet OR smart contract bytecode.

02

Run checks

Agent runs on-chain analytics + sanctions screening + regulatory rule matching.

03

Jurisdiction match

Cross-references against your declared operating jurisdictions.

04

Return verdict

Returns verdict (allow / review / block) + specific regulations triggered + remediation.

Sandbox

Try it live

Simulate compliance evaluation across jurisdictions, chains, and counterparty types.

Preview — same deterministic engine as the API, no key required.See API pricing
Inputs
Evaluation type
Chain
Amount (USD equivalent)
Originator jurisdiction
Beneficiary jurisdiction
Counterparty type
Response
{
  "verdict": "allow",
  "regulations_triggered": [
    {
      "framework": "FATF Travel Rule",
      "article": "Recommendation 16",
      "severity": "info",
      "action_required": "Exchange originator and beneficiary information between VASPs.",
      "jurisdiction": "EU"
    }
  ],
  "on_chain_risk_score": 15,
  "sanctions_screening": "clear",
  "evaluated_at": "2025-05-25T12:00:00.000Z"
}

Simulated response — no real inference is performed.

Data sources

Grounded in public regulatory frameworks

  • Live: US Treasury OFAC SDN list — ingested and refreshed daily, screened per request
  • Live: EU consolidated financial sanctions list — ingested and refreshed daily
  • Live: Etherscan public data (Ethereum, Polygon) for contract verification and address age — no clustering or mixer tracing
  • Encoded frameworks: FATF Recommendation 16 (Travel Rule), EU MiCA, US FinCEN thresholds
  • Planned — not yet connected: UK OFSI consolidated list (pending a stable published URL), Bitcoin/Tron explorer coverage, and paid chain-analytics attribution

All sources are public and updated as guidance evolves.

Smart contracts

Contract address checks

When evaluation_type is set to smart_contract, the agent screens the contract address against the live sanctions lists and reports its public verification status and age. It does not perform static vulnerability analysis, and says so in the response.

What the contract check returns

Public explorer data plus sanctions screening — limitations included in every response

Source verification status on the public explorer
Contract age (first-seen timestamp) as a deployment-risk signal
Sanctions screening of the contract address against OFAC and EU lists
Planned — not yet available: static bytecode/source vulnerability analysis (SWC classes)

Which frameworks does this agent check?

+

FATF Travel Rule, MiCA, and OFAC sanctions screening, alongside common AML and KYT signals. Rules are grounded in publicly published regulatory text, so verdicts cite the specific clause or sanctions list that triggered each finding.

Can it assess smart-contract risk in the same API call?

+

Yes. The same evaluate call can score counterparty contract risk — flagged deployers, mixers, known exploit patterns — alongside the transaction-level Travel Rule and OFAC checks, so you get one structured verdict per request.

How quickly can I ship compliance checks with this agent?

+

Most teams ship in a day. You POST encrypted transaction metadata to /v1/agents/crypto-compliance/evaluate, get back a structured JSON verdict with per-rule status and remediation, and wire that into your existing compliance or transaction-blocking flow.

Does screening happen without exposing the underlying transaction data?

+

Yes. The agent operates on encryption metadata and the screening attributes you choose to send — addresses, amounts, jurisdictions — not the underlying transaction body. Sensitive payload data stays encrypted under keys you control.

Is there a free sandbox to evaluate it?

+

Yes. The interactive sandbox lets you submit synthetic transactions and see FATF, MiCA, and OFAC verdicts live. Validate the response shape and rule coverage before requesting a production API key or wiring it into transaction authorization.

Ready to integrate?

Get an API key and start calling AI Crypto Compliance Agent in minutes.