Methodology

anonym.legal makes specific claims about detecting and anonymizing personal data. This page explains how the system actually works, how we arrive at the numbers we publish, and — just as importantly — where the method has limits, so you can judge it against your own compliance standard.

How detection works

Detection combines machine-learning named-entity recognition with deterministic pattern matching and contextual validation. The model proposes spans of likely personal data; pattern rules confirm structured identifiers (emails, card numbers, national IDs) by format and checksum where one exists; and surrounding context is used to resolve ambiguous mentions. A span is only redacted when these signals agree, which keeps over-redaction of ordinary text low without relying on a single fragile heuristic.

Entity coverage and how we count it

We publish coverage of 285+ entity types. That figure counts the distinct categories the system is configured to recognize — names, contact details, government identifiers, financial data, health data, and domain-specific identifiers — not the number of patterns or model labels behind them. Coverage varies by entity: well-structured identifiers are detected more reliably than free-form references, and you can narrow or extend the active set per job using presets.

Language support

Detection runs across 48 languages. Accuracy is highest for languages with large training corpora and regular orthography, and lower for low-resource languages, transliterated text, and mixed-language documents. We do not claim identical performance across every language, and we recommend reviewing output on a representative sample before processing a new language at volume.

Reversible vs. irreversible anonymization

You choose the transformation. Irreversible redaction removes the original value and cannot be undone. Reversible tokenization replaces values with consistent tokens that an authorized holder of the key can later restore — useful when an auditor or downstream system must verify figures against source records, then have access revoked. Reversible mode is pseudonymization under the GDPR, not anonymization; treat tokenized output as still personal data while the mapping exists.

Accuracy, false positives, and false negatives

No automated system detects personal data perfectly. False negatives (missed PII) and false positives (redacting non-PII) both occur, and their rates depend on document type, language, formatting, and OCR quality on scanned material. We tune toward recall on sensitive categories, but residual risk remains. anonym.legal is a control that reduces exposure — it is not a guarantee of complete de-identification.

Human verification and your responsibility

Because detection is probabilistic, a human should review output before sensitive material is released, especially for legal, healthcare, and regulatory use. You remain the controller of your data and are responsible for validating that anonymization meets your legal standard — for example HIPAA Expert Determination or a GDPR anonymization assessment. We document limitations openly so you can build the right review step around the tool rather than assume it away.

Questions about the method

Want detail on how a specific entity type or language is handled, or how a published number is derived? Ask — we would rather explain than overstate.

About this page

We update this page when our platform or the law changes.

Read our founder note for how we work.

Each change shows up in the timestamp at the top.

We follow these rules

GDPR (EU 2016/679).
ISO/IEC 27001:2022.
NIS2 (EU 2022/2555).
HIPAA safe harbor under 45 CFR § 164.514(b)(2).

Our promise

We do not sell your data.

We do not train models on your text.

We store your files in Germany.

You can delete your account at any time.

You own your work.

Where we run

Our servers live in Falkenstein, Germany.

We use Hetzner. They hold ISO 27001 certification.

All data stays in the EU.

Backups run every day.

Need help?

Email support@anonym.legal.

We reply within one business day.

How we test

We run a full check suite on every release.

Each surface gets its own sweep script and report.

Human reviewers spot-check the output each week.

We track recall and precision on a labelled set.

Bad runs block the deploy.

What we never do

We never sell your information to third parties.
We never train models on what you upload.
We never keep your work after you delete it.
We never share keys with any outside firm.
We never run ads inside the product.

Plans in plain words

We sell credits, not seats.

One credit covers one short job.

Long jobs use a few credits each.

You can top up at any time.

Unused credits roll over each month.

Read the plans page for current rates.

Who built this

A small team of engineers and lawyers built this.

We ship from Europe and work in the open.

Our founder note spells out why we started.

Where to start

How the parts fit

A browser add-on cleans text inside Chrome.

A Word plug-in handles drafts in Office.

A small desktop tool works on whole folders.

An agent protocol link feeds large models safely.

All four share one core engine and one rule set.

Words from our team

We started this work after a lunch about cookies.

One friend kept getting odd ads on her phone.

We asked why a court file leaked through a draft.

We sketched the first build on a napkin that week.

By month three we had a tiny demo for a friend.

She used it on her first case the next day.

Common questions we hear

Can the tool read scanned PDFs? Yes, with OCR.

Does it work on long files? Yes, in small chunks.

Can I roll my own rule set? Yes, save it as a preset.

Does it run offline? The desktop build runs offline.

Do you keep my files? No, the cloud build wipes after each run.

Will it learn from my work? No, we never train on inputs.

A short tour of the workflow

Upload a file or paste a snippet of prose.

Pick the entities you want gone from the draft.

Choose a method: replace, mask, hash, encrypt, or redact.

Press run and watch the side panel show each hit.

Skim the result and tweak any rule that misfired.

Save the cleaned file or send it to a teammate.