FinTechReview Prompt

Review Prompt to Map UK CDD Evidence Against Every Reading of MLR 28 with Luna

This **Luna** review prompt maps your UK CDD evidence against the competing readings of **Money Laundering Regulation 28** while the guidance is unsettled. Luna, the deepidv compliance agent, runs a **method census** of every verification path, a **reading-by-reading compliance** check, a **weakest-file review** of your thinnest ten records, an **enforcement-readiness** file-request procedure for the £500M investigator build-out, and a **re-mapping plan** for the day the guidance settles.

How to use this prompt

1
Open Luna in the deepidv dashboard and paste the full prompt, or run it in a general LLM to draft the committee brief outside the platform.
2
Confirm the customer book scope and the verification paths you actually run.
3
Review the re-verification exposure Luna quantifies per reading and prioritize the weakest files.
4
Adopt the standing file-request response procedure so an investigator's letter is a repeat performance.

The prompt

Luna, map our CDD evidence against the competing readings of UK Money
Laundering Regulation 28, given the current guidance conflict: OfDIA
points to DVS-certified providers, the FCA is silent, and Regulation 28
also recognizes EU eIDAS qualified trust services.

For our UK customer book:
1. Method census: classify every CDD verification by path, DVS-certified
   credential, eIDAS qualified credential, chip/forensic document
   verification with liveness, or data-source match only, with volumes
   per path.
2. Reading-by-reading compliance: for each live interpretation of MLR 28
   (DVS-required, eIDAS-sufficient, evidence-based), mark which paths
   satisfy it and which customers would need re-verification if that
   reading is confirmed.
3. Weakest-file review: pull the ten customer files with the thinnest
   verification evidence and assemble their complete CDD records,
   identification, screening timestamps, risk rationale, monitoring
   dispositions, timing how long assembly takes.
4. Enforcement readiness: given the £500M strategy's investigator
   build-out, draft the standing file-request response procedure, who
   produces what, from which system, in what timeframe.
5. Re-mapping plan: the configuration changes required the day guidance
   settles, in either direction, with no customer-facing rebuild.

Deliver as a compliance committee brief with the re-verification
exposure quantified per reading.

Test it in Claude or another LLM

This prompt is built for the Luna agent inside deepidv, where Luna reads your live CDD records and screening history. You can dry-run the mapping logic in a general LLM with a synthetic method census before running it on the real book.

1
Paste the full prompt into Claude, ChatGPT, or Gemini, replacing Luna, with a role instruction such as: You are a UK AML compliance analyst.
2
Under the customer-book section, paste the synthetic method census below.
3
Ask the model for the reading-by-reading compliance table and the re-verification exposure per reading.

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.

CUSTOMER BOOK (synthetic): 120,000 UK retail customers. Verification paths: 60 percent chip/forensic document with liveness, 25 percent data-source match only, 10 percent DVS-certified credential, 5 percent eIDAS credential.

Pairs with on deepidv

Sources & further reading

FAQ

What does the MLR 28 Evidence Mapper produce?

A compliance committee brief with the re-verification exposure quantified per reading of Regulation 28: a method census, a reading-by-reading compliance check, a weakest-file review, an enforcement-readiness procedure, and a re-mapping plan for when the guidance settles.

Why map against multiple readings of MLR 28?

Because the guidance is in conflict: OfDIA points to DVS-certified providers, the FCA is silent, and Regulation 28 also recognizes eIDAS qualified services. Verifying to the strictest coherent reading now is cheaper than re-verifying a customer book later.

Can Luna time how long a CDD file takes to assemble?

Yes. The weakest-file review pulls your thinnest ten records and assembles their complete CDD files, identification, screening timestamps, risk rationale, and monitoring dispositions, timing the assembly so you know your answer speed before an investigator asks.

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