FinTechReview Prompt

Compliance Prompt for FinCEN BSA Modernization Audits

This prompt turns Luna, the deepidv regulatory intelligence agent, into a FinCEN modernization auditor. It maps your corporate onboarding pathways against the proposed results-driven compliance criteria, ranks manual bottlenecks by review dwell time, and proposes agentic configurations plus formal comment-letter language. Built for fintech compliance leads and BSA officers preparing for FinCEN's shift from checkbox compliance to measured effectiveness.

How to use this prompt

1
Paste the prompt into Luna in your deepidv dashboard, or run it in Claude, ChatGPT, or Gemini if you want a standalone point-in-time audit without live regulator feeds.
2
Replace the input section with your onboarding pathway inventory, effectiveness metrics, manual review queue volumes, and any prior FinCEN, FDIC, or state-regulator findings.
3
Expect a five-part output: a results-driven gap map, a manual bottleneck inventory, an agentic configuration proposal, an effectiveness metric set, and draft comment language.
4
Route the gap map to your BSA officer and the draft comment language to outside counsel before filing with FinCEN.
5
Re-run the loop monthly so the gap map stays current with each new FinCEN, FATF, or state-regulator publication.

The prompt

Luna, evaluate our current corporate onboarding pathways against FinCEN's proposed results-driven compliance criteria ahead of the June 9 comment deadline. Audit our processing tracks for manual bottlenecks and suggest automated agentic configurations to ensure continuous tracking compliance.

INPUT, the user will paste:
- Corporate onboarding pathway inventory and the systems each one touches
- Current effectiveness metrics (detection rate, time-to-disposition, false-positive rate by detection scenario)
- Manual review queue volume and the average dwell time per case
- Any prior FinCEN, FDIC, or state-regulator findings

OUTPUT, return the following structured response:

1. RESULTS-DRIVEN GAP MAP
For each FinCEN modernization criterion:
- The criterion and the section reference
- Current firm-level coverage (covered, partial, gap)
- Specific evidence of coverage or the evidence-absence reason

2. MANUAL BOTTLENECK INVENTORY
- Each onboarding step ranked by manual-review dwell time
- The specific data dependency causing the bottleneck
- The recommended automated replacement and the agent that owns it

3. AGENTIC CONFIGURATION PROPOSAL
- Specific Luna rule sets to deploy against FinCEN, FATF, and state-regulator publication feeds
- Specific Arbiter red-team scenarios to schedule
- Specific Arc gateway integrations to wire for credential ingestion

4. EFFECTIVENESS METRIC SET
- The metric set the firm should publish internally before the comment window closes
- The cadence (daily, weekly, monthly) and the owner per metric
- The escalation path when a metric drifts outside the agreed threshold

5. COMMENT SUBMISSION
- Recommended language for the firm's formal comment to FinCEN before June 9
- The specific proposed-rule sections the firm should address
- Open questions for outside counsel review

Be specific. Cite section references where you can. Where input is insufficient to assess a criterion, flag the question instead of guessing.

Test it in Claude or another LLM

This prompt is built for the Luna agent inside deepidv, where it audits your corporate onboarding pathways against FinCEN's proposed results-driven compliance criteria and proposes agentic configurations ahead of the comment deadline. Here is how to dry-run the same workflow in any general LLM with synthetic data before you wire it to live deepidv onboarding and metric feeds.

1
Paste the full prompt into Claude, ChatGPT, or Gemini, and replace the direct address 'Luna,' at the start with a role instruction such as 'Act as a BSA/AML compliance program lead reviewing our onboarding against FinCEN modernization criteria.' Keep the five OUTPUT sections (gap map, bottleneck inventory, agentic config, metric set, comment submission) exactly as written.
2
Below the prompt, paste the synthetic sample data block from sampleInput so the LLM has a fake onboarding inventory, effectiveness metrics, manual queue stats, and a prior finding to reason over. This stands in for the live deepidv onboarding and metrics export.
3
Add one line telling the model to treat any criterion it cannot assess from the fake data as a flagged open question rather than inventing coverage, mirroring the prompt's 'flag the question instead of guessing' instruction.
4
Good output for this prompt is a per-criterion gap map labeling each item covered/partial/gap with a section reference, a bottleneck table ranked by dwell time with a named automated replacement and owning agent, a concrete metric set with cadence and escalation path, and draft comment language naming specific proposed-rule sections. If the model returns vague prose without the five numbered sections or the covered/partial/gap labels, tighten the role line and re-run.
5
Once the output shape and specificity look right, run the prompt live in the deepidv dashboard where Luna executes it against your real onboarding pathways, effectiveness metrics, and regulator-finding history.

Synthetic sample data to paste alongside the prompt

Fake test data, safe to share with any LLM. Swap in your own once the output looks right.

ONBOARDING PATHWAY INVENTORY (synthetic, fake):
- Track CORP-TEST-01 (LLC fast lane): touches ACME-IDV-SANDBOX, RegistryMock API, ManualQueue-A
- Track CORP-TEST-02 (trust/nominee): touches ACME-IDV-SANDBOX, ManualQueue-A, ManualQueue-B
EFFECTIVENESS METRICS (fake): detection rate 71%, time-to-disposition 5.2 days, false-positive rate 18% on UBO-mismatch scenario
MANUAL QUEUE: ManualQueue-A volume 340 cases/wk, avg dwell 41 hrs (UBO verification); ManualQueue-B volume 90 cases/wk, avg dwell 63 hrs (sanctions adjudication)
PRIOR FINDING (fake): State-Reg-XX 2025 exam noted inconsistent EDD documentation on nominee structures, ref FINDING-TEST-0000

Pairs with on deepidv

Sources & further reading

FAQ

What is FinCEN's results-driven compliance modernization?

FinCEN has proposed shifting Bank Secrecy Act supervision away from checkbox-style technical compliance toward measured program effectiveness, asking institutions to demonstrate outcomes such as detection rates, time-to-disposition, and false-positive performance. This prompt audits your corporate onboarding against that posture and drafts the internal metric set and comment-letter language needed to respond.

Can I run this FinCEN audit prompt outside the deepidv dashboard?

Yes. The prompt is written for Luna, the deepidv regulatory intelligence agent, but it works in Claude, ChatGPT, or Gemini if you paste your pathway inventory and metrics into the input section. You lose live regulator-feed monitoring that way, so treat external runs as point-in-time audits rather than continuous tracking.

Run it with live verification data

These prompts work in any LLM. Inside the deepidv dashboard, Luna, Arbiter, and Arc run them against your real sessions, screening lists, and audit trails.

Book a Demo