Tax advisors · Workflow 04
Clarify an ownership chain before preparing a Transparenzregister filing.
Ask an approved AI tool to organise the ownership facts and flag information still needed for professional assessment. Do not ask it to submit the filing or declare the beneficial owner from an incomplete chart.
The moment this is useful
The task on your desk
A client asks your tax advisory team to help prepare information for the German Transparency Register. The shareholder chart includes a holding company, two natural persons, and a note about separate voting arrangements. Before requesting missing documents, you want to turn the chart into a clear list of questions for the responsible reviewer.
Keep the facts the question depends on.
Keep for the research
- Every relevant ownership percentage and the direction of each holding.
- Voting rights and other control arrangements, separately from capital interests.
- Entity types and the German regulatory context needed for research.
Remove or replace
- Client, shareholder, and family names; addresses and dates of birth.
- Register identifiers, tax numbers, document links, and contact information.
- Personal background unrelated to understanding the ownership or control chain.
A worked prompt example
Fictional details · Edited by hand after reviewBefore review
Elbe Services GmbH is owned 60% by Hafen Holding GmbH and 40% by Nora Beck. Hafen Holding is owned 50/50 by Nora Beck and Paul Stein. There may be a separate voting agreement. Please prepare questions before we finalise the Transparenzregister information.
Reviewed prompt
Hypothetical German structure: Company A is owned 60% by Holding B and 40% by Person 1. Holding B is owned 50/50 by Person 1 and Person 2. A separate voting agreement may exist but has not been reviewed. Organise the known ownership facts and list the documents and questions needed before assessing beneficial ownership for the Transparenzregister. Consider ownership, voting rights, and other control separately. Do not infer a final beneficial-owner determination from arithmetic alone.
Use this workflow
- Convert the relevant chart into a short text description; keep the source documents in the client file.
- Detect personal information locally and review each highlighted identity.
- Manually assign consistent company/person aliases and verify the chain, percentages, and voting-right notes.
- Use the output to request missing documents. Check the legal assessment and actual filing data separately against official guidance.
What you should have at the end
A targeted request for the missing voting agreement and ownership evidence, plus a checklist for the professional reviewing the filing.
Redacted research text is not filing data. An actual register submission needs the required real information and a verified assessment of the applicable obligations.
Questions for this workflow
Should ownership percentages be removed?
They are central to this research question. If your approved workflow permits sharing the structure, retain the percentages and relationships while replacing identities. If the structure itself is confidential or distinctive, use a hypothetical or an approved private environment.
Can I use the generated aliases in the filing?
No. Aliases are for the research prompt only. Prepare the actual filing from verified source information and the register’s current requirements.
Workflow context & further reading
These sources inform the workflow. The example and editing choices are Procurato’s illustration; they are not source-endorsed advice.
- GwG § 3 — beneficial owner — The statutory starting point for ownership and control questions.
- GwG § 19 — particulars of the beneficial owner — Why actual filing information must be distinguished from a redacted research prompt.
- Official Transparenzregister — The register’s official entry point for filing information and guidance.
From example to your own work
Keep the context. Review the details.
Paste into the local redactor, review each detection, choose what to keep, and check the result before sharing it with your approved AI tool. Names are only part of the review: inspect confidential wording and identifying context too.
Free, no account. Initial model download: about 1.1 GB. Requires WebGPU and several GB of free memory. Detection focuses on English and can miss details. The examples here include manual editing; the tool does not automatically rewrite or summarise your text.