Interpreter Transcript Redaction with anonym.plus

Clear identifiers from both language tracks of an interpreted record.

In simple terms, PII redaction is the on-device process of finding and masking personally identifiable information in a document before it is shared.

Interpreter transcript redaction is the removal of personal identifiers from a record taken with an interpreter. It supports CPR 5.4C and the DPA 2018 before disclosure. anonym.plus reads 48 languages, so it flags IDs in both the source and the rendered text.

When this applies

A deposition uses an interpreter, so the record holds two language tracks. Names and numbers appear in both, so each must be hidden before filing.

How anonym.plus handles it

  1. Load the record into anonym.plus on your device.
  2. Set the source language so both tracks are read.
  3. Names, dates, and numbers get flagged in each track.
  4. Swap or black out the confirmed identifiers.
  5. Save the clean record on your device.

What you need to provide

PII entity types detected

Categoryanonym.plus entity typeExample
NamesPERSONdeponent → [DEPONENT]
NamesPERSONthe interpreter → [INTERPRETER]
DatesDATE_TIMEdate → [DATE]
IdentifiersNATIONAL_IDnational ID → [ID]
LocationLOCATIONaddress → [ADDRESS]
CaseCASE_NUMBERclaim no. → [CASE_NO]

Compliance achieved

Anonymise interpreter transcripts offline — see plans & start free →

Limitations & cautions

ID and date formats differ by language, so set the source tongue correctly. A name rendered two ways across tracks may need a shared label. Review both tracks before you export.

Frequently asked questions

How many languages does the tool read?

It reads 48 languages. That matters where the evidence is taken abroad on a letter of request under CPR 34.13, or taken here for a foreign court under s.2(2)(a) of the Evidence (Proceedings in Other Jurisdictions) Act 1975, because the source and rendered tracks use different identifier and date formats.

Are both language tracks cleaned?

Yes. IDs are flagged in the source text and the rendered text, so neither track names a real person.

Is the interpreter named too?

The interpreter's name is personal data and can be flagged. You choose to keep or swap it.