AUM dataset anonymisation is the removal of personal data from an assets-under-management table so that the remaining rows are no longer about identifiable people. UK GDPR Recital 26 draws that line: the law does not apply to anonymous information, and the test is whether anyone could re-identify the person by means reasonably likely to be used — the ICO's motivated-intruder test. Art. 4(5) makes the opposite point: pseudonymised data that can be reversed with extra information is still personal data. DPA 2018 s.3(2) supplies the statutory definition. anonym.plus marks each identifier on your device, so the figures stay while re-identification becomes impractical.
When this applies
An assets table pairs client identities with balances, mandates and regions for analytics. A modelling team, a board pack or a vendor benchmark needs the distribution, not the clients. You strip the direct identifiers and generalise the quasi-identifiers so Recital 26 treats the rows as anonymous.
How anonym.plus handles it
- Open the table in anonym.plus on your device.
- The tool flags names, National Insurance numbers, IBANs and contact columns.
- Turn the reversible alias map OFF — a kept map means Art. 4(5), not anonymity.
- Keep the balance, mandate and segment columns.
- Replace each direct identifier, then band the standout amounts.
- Save the clean table locally.
What you need to provide
- The assets table (CSV, XLSX, or DOCX).
- The Replace operator with the alias map turned off.
- Optional allow-list for segment, mandate and region codes.
PII & financial identifiers detected
| Category | anonym.plus entity type | Example |
|---|---|---|
| Names | PERSON | client name → [REMOVED] |
| Identifiers | UK_NINO | NINO → [REMOVED] |
| Financial | IBAN_CODE | IBAN → [REMOVED] |
| Money | MONEY | funds £4.1M → [VALUE] |
| Contact | EMAIL_ADDRESS | client email → [REMOVED] |
| Location | LOCATION | town → [REGION] |
Compliance achieved
- UK GDPR Recital 26 puts truly anonymous information outside the regulation entirely, judged on whether re-identification is possible by means reasonably likely to be used — the ICO frames this as the motivated-intruder test.
- UK GDPR Art. 4(5) is the trap: data that can be re-attributed using separately held information is pseudonymised, not anonymous, and stays fully in scope. Turning the alias map off is what makes the difference.
- Where the table is used for statistical or research purposes, Art. 89(1) requires safeguards and data minimisation, with anonymisation preferred wherever the purpose can still be met.
- DPA 2018 s.3(2) supplies the statutory definition of personal data; the whole assessment runs offline, so the raw table never reaches a server.
Anonymise AUM datasets offline — see plans & start free →
Limitations & cautions
Recital 26 counts indirect re-identification, so a rare balance plus a region can still single out one client after every name is gone. The tool removes direct identifiers; it does not decide whether the residual risk is acceptable. Generalise quasi-identifiers — band amounts, widen regions — and document the assessment.
Frequently asked questions
When is data anonymous under UK GDPR Recital 26?
When no one can re-identify a person by means reasonably likely to be used, taking account of cost, time and available technology. The ICO's motivated-intruder test is the practical version. Removing direct identifiers is only the first step; quasi-identifiers usually need generalising too.
Why turn the alias map off?
Because a stored map is exactly the 'additional information' in Art. 4(5). With it, the dataset is pseudonymised and still in scope; without it, you have a genuine shot at the Recital 26 standard.
Is the raw table uploaded?
No. The application is fully offline, so the dataset never leaves your machine at any point in the assessment.