Batch redaction is the removal of PII from a whole folder of agreements in one run. A contract library is held for years because Limitation Act 1980 s.5 allows six years to sue on a simple contract, and s.8 allows twelve on a deed. UK GDPR Art. 5(1)(e) still asks you to keep personal data no longer than you need it. Redacting the working copies is how both can be true, and a batch handles up to 20 files.
When this applies
A precedent library or an audit means cleaning many agreements at once, with the same party in several. The ICO's anonymisation guidance applies a motivated-intruder test: could a reasonably competent person, using public sources, put a name back? Across a set, small clues combine, so one consistent pass matters more than it does on a single file.
How anonym.plus handles it
- Point anonym.plus at the folder on your machine.
- It scans each file for the full PII set.
- A shared map keeps repeat parties steady across the run.
- Review the summary and fix any low-confidence flags.
- Swap or black out the confirmed IDs.
- Save the clean library on your device.
What you need to provide
- A folder of agreements (PDF, DOCX, or mixed; up to 20 files).
- The shared label map turned on for steady results.
- An operator (Replace works well for a library).
PII entity types detected
| Category | anonym.plus entity type | Example |
|---|---|---|
| Parties | ORGANIZATION | party across files → [PARTY_1] |
| Names | PERSON | repeat signatory → [SIGNATORY_1] |
| Dates | DATE_TIME | deal dates → [DATE] |
| Banking | IBAN_CODE | accounts → [IBAN_n] |
| Location | LOCATION | addresses → [ADDRESS] |
| Contact | EMAIL_ADDRESS | emails → [EMAIL] |
Compliance achieved
- Applies one standard to every file, which is what the ICO's motivated intruder test asks of a set.
- Supports retention judged against Limitation Act 1980 s.5 (six years) and s.8 (twelve years on a deed).
- Serves the storage-limitation principle in UK GDPR Art. 5(1)(e) for working copies.
- The shared map keeps results steady across the run.
- Whole-folder work stays offline — nothing uploaded.
Anonymise contract portfolios offline — see plans & start free →
Limitations & cautions
A batch handles up to 20 files at a time, so split a large library into runs. Mixed folders lean on OCR for scans, so review low-confidence flags. Guard the shared map, since it can re-link the set if kept.
Frequently asked questions
What is the motivated intruder test?
It is the yardstick the ICO's anonymisation guidance uses: would a reasonably competent person with no special skills, using public sources, be able to identify someone from the released data? It is deliberately more demanding than asking whether a casual reader could. Across a batch it matters more, because clues in different files combine.
How many files can one batch take?
Up to 20 at a time, so split a larger library into several runs. Each run keeps its own steady labels across the files it covers.
How does batch mode keep one party steady?
A shared map logs each party once, so the same firm or signatory maps to one alias in every file across the folder. Guard that map: while it exists the set is pseudonymous, not anonymous.