FinTechReview Prompt

Review Prompt to Audit Your Corporate Book Against UK Director ID Deadlines

This **Luna** review prompt treats the September 2026 UK convictions as the enforcement template, where person-level verification failures are prosecutable and a verified director is liable for an unverified colleague, and audits your entire corporate client book against it. Luna, the deepidv compliance agent, builds a **board verification matrix** for every director and PSC with status, method, date, and evidence reference, flagging any 'verification' that is only a registry lookup rather than liveness-anchored identity proofing. It maps **deadline exposure** against the November 2026 transition window and each client's next confirmation-statement date, produces the **cross-liability view** pairing verified directors with their unverified colleagues, runs a **nominee screen** for recurring faces and sub-200-day entities moving unexplained volume, and lists **evidence gaps** where a complete per-director record could not be produced within 24 hours. Built for compliance and onboarding leads who now have to defend business verification in a courtroom, not a filing cabinet.

How to use this prompt

1
Open Luna in the deepidv dashboard and paste the full prompt, or run it in Claude, ChatGPT, or Gemini if you are drafting the exposure register outside the platform.
2
Replace the INPUT section with your corporate client list, each board's directors and PSCs, and the current verification status and method per person.
3
Run the prompt and read the board verification matrix first: it separates genuine liveness proofing from registry lookups treated as verification.
4
Work the exposure register top down, since it is ranked by days-to-deadline and cross-liability, and assign an owner and remediation date to every gap.
5
Re-run monthly, and immediately after any board change, so the register stays current as the November 2026 window closes.

The prompt

Luna, audit our corporate client book against the UK director identity verification regime, treating the September 2026 convictions as the enforcement template: person-level failures are prosecutable, and a verified director is liable for an unverified colleague.

Across all corporate accounts:
1. Board verification matrix: every director and PSC per client, with verification status, method, date, and evidence record reference. Flag any person whose "verification" is a registry lookup rather than identity proofing with liveness.
2. Deadline exposure: clients with unverified directors mapped against the November 2026 transition window and each client's next confirmation-statement date, ranked by days remaining.
3. Cross-liability view: for each verified director, list the unverified colleagues on the same board, since that pairing is the exact fact pattern the first prosecution punished.
4. Nominee screen: directors or faces recurring across unrelated clients, verification sessions inconsistent with declared roles, and sub-200-day entities moving volumes their history cannot explain.
5. Evidence gaps: accounts where we could not produce a per-decision verification record for every director within 24 hours of a regulator's request, with owner and remediation date per gap.

Output as a ranked exposure register: client, gap, deadline, owner, date. Treat "the registry accepted the filing" as no evidence.

Test it in Claude or another LLM

This prompt is built for Luna inside deepidv, where Luna reads live client records and per-decision evidence. You can dry-run the audit shape in any general LLM first with synthetic board data before wiring it to real accounts.

1
Paste the full prompt into Claude, ChatGPT, or Gemini, replacing the opening 'Luna,' with a role instruction such as 'Act as a corporate verification auditor mapping a client book against UK director ID deadlines.' Keep the OUTPUT sections as written.
2
Under the INPUT section, paste the synthetic sample block below so the model has boards and verification statuses to audit.
3
Add one framing line: 'This is synthetic test data. Where a verification method is unspecified, treat it as no evidence rather than assuming liveness.'
4
Check the output shape: a board matrix, a deadline ranking, a cross-liability view, a nominee screen, and an evidence-gap register. If any entry claims liveness the input did not state, tighten the framing and re-run.
5
Once the shape is right, run it live in the deepidv dashboard where Luna pulls the real board and evidence records.

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.

CLIENTS (synthetic, fake):
- Acme Ltd: directors Jane Roe (verified via liveness 2026-03), John Doe (registry lookup only), next confirmation statement 2026-10-15
- Zed Holdings: director A. Nominee (status unknown), entity age 140 days, volume inconsistent with trading history
EVIDENCE (fake): no per-decision record retained for John Doe
OPEN ITEM (fake): whether registry acceptance of Zed's filing counts as verification (it does not)

Pairs with on deepidv

Sources & further reading

FAQ

What are the UK director identity verification deadlines?

Newly appointed directors must verify their identity with Companies House before acting, and existing directors must verify at their next confirmation statement inside a transition window that closes in November 2026. The first convictions for failures landed in September 2026.

What does this prompt produce?

A board verification matrix per client, a deadline-exposure ranking, a cross-liability view pairing verified directors with unverified colleagues, a nominee screen, and an evidence-gap list, output as a single ranked exposure register with an owner and date per gap.

Can I run this outside the deepidv dashboard?

Yes. The structure works in Claude, ChatGPT, or Gemini as an audit framework and returns the matrix and register from data you paste. Live board data, evidence-record lookups, and 24-hour production only run when it executes inside the deepidv dashboard through Luna.

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