Conduct spine · Obligation 07

    Financial-Crime Controls.

    AML / BSA programme assurance. Attests that an issuer's financial-crime programme is functionally equivalent in outcome to a bank-grade BSA programme — the control set the banks say is missing and the dissenting senators say is absent.

    Phase C1 (core substantial-similarity engine) · Phase C2 (full programme + AI Assurance OS bindings)

    Legislative status · verified 26 July 2026

    H.R.3633 (Digital Asset Market Clarity Act) passed the House 294–134 on 17 July 2025 and sits on the Senate Legislative Calendar (General Orders, Calendar No. 423) after the Banking Committee reported a substitute on 1 June 2026. Updated merged Banking/Agriculture text was released on 22 July 2026. Senate leadership expects the pre-recess floor window to be missed, and the draft still requires 60 votes. Cabier plans on the operative regimes — state trust charters and money-transmitter licensing, OCC, FRB and FDIC prudential guidance, and EU MiCA Titles III and IV for cross-border cohorts — binding through at least 2027. Legislation is the label on the control, not the control.

    Source: congress.gov, H.R.3633 (119th Congress) — legislative actions and Senate Legislative Calendar, General Orders No. 423.

    What governs tokenised dollars now

    Framing

    Position is enforcement-equivalence attestation — not "AML is absent". CLARITY §110 already designates digital-commodity brokers, dealers and exchanges as BSA financial institutions, and GENIUS already brings payment stablecoins under the BSA. The residual is operating-conduct attestation that the programme works in outcome, and that it applies equally to a bank issuer and a digital-asset issuer.

    Sub-controls (eight)

    1. FR-07-01

      Customer due diligence & KYC onboarding

      CIP, CDD, beneficial-ownership identification, and enhanced due diligence for high-risk relationships, with evidence-graded coverage.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    2. FR-07-02
      AI Assurance OS binding

      Customer risk-rating model

      Attested via the AI Assurance OS — model registry, SR 11-7 workflow, bias audit.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    3. FR-07-03
      AI Assurance OS binding

      Transaction-monitoring coverage & rule effectiveness

      Scenario coverage, tuning, and alert-disposition quality; monitoring models registered in the AI Assurance OS.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    4. FR-07-04

      SAR / STR generation, quality & timeliness

      Suspicious-activity reporting workflow, decision quality, and filing-timeliness attestation.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    5. FR-07-05
      Cross-link 06

      Sanctions / OFAC & list screening

      Screening coverage and tuning; cross-linked to Obligation 06 (Cross-border conduct) as the shared screening evidence source — no duplicate store.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    6. FR-07-06

      Travel Rule (FATF R.16)

      Originator / beneficiary data transmission and counterparty-VASP handling.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    7. FR-07-07

      Recordkeeping & retention

      Retention, retrievability, and audit-trail integrity for financial-crime records.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    8. FR-07-08

      Independent programme effectiveness review

      Human-led 3LOD review producing the Substantial-Similarity Attestation and remediation artefacts.

      grade: ▒▒ · evidence: ▒▒ · trust gate: ▒▒

    Named outputs

    • · AML Programme Effectiveness Grade
    • · Substantial-Similarity Attestation
    • · Remediation artefacts (3LOD-reviewed)

    Regulatory anchors

    • · Bank Secrecy Act (BSA) [verify]
    • · FinCEN CDD Rule [verify]
    • · AMLA 2020 [verify]
    • · FATF Recommendations 10, 16, 20 [verify]
    • · OFAC [verify]
    • · GENIUS Act — BSA application to payment stablecoins [verify]
    • · CLARITY Act §110 — digital-commodity broker/dealer/exchange as financial institution [verify]

    Acceptance

    Each sub-control yields a graded outcome with linked evidence. Obligation produces a single named attestation. All model-driven sub-controls show a live AI Assurance OS binding. No methodology is exposed in any output.