Anonymize documents automatically with AI

Comply with GDPR before publishing documentation or sending it to AI.

◇ Works with .docx · .pdf · .odt · Data sovereignty

No credit card

Try anonimizia free for 14 days

Anonymize documents with no page limit. No card required. We'll reply within 24 h with access configured for your organization.

  • No credit card
  • 24 h support
  • Cancel anytime

By submitting you accept our privacy policy. No spam.

Already have an account? Sign in
The process

Four steps. Zero exposure.

01

Upload

Arrives via API from your DMS, or drag a PDF, Word or a batch of documents.

02

Detect

The AI locates names, IDs, addresses and 25+ personal data types.

03

Review

Pick which categories to redact, adjust each entity manually and confirm before exporting.

04

Publish

Anonymous document ready, with anonymization report and full audit trail.

The product

Built for public documents, not generic ones.

anonimizia understands how a file, tender or decree is structured. And that changes everything.

Configurable & multi-layer

Choose which categories to anonymize per document type and let models trained on Iberian administrative vocabulary detect national IDs, IBAN, license plates and cadastral references.

Anonymization reports with review

Review panel with confidence level per entity: accept, edit or discard each redaction. Every document generates a report of what was redacted, why and who reviewed it.

Integrable and/or cloud

Connect it via REST API to leading public-sector document management systems, or run it in our sovereign EU cloud.

Anonymization modes

Anonymize or pseudonymize: you choose how.

Anonymizing removes the data irreversibly and takes the document out of GDPR scope. Pseudonymizing replaces it with an equivalent that keeps the document readable and usable, with controlled re-identification. anonimizia does both, field by field.

Anonymize (irreversible)

The personal data disappears from the document and from its text layer: it cannot be recovered from the published file.

Redaction (black bars)

An opaque bar covers the value and the underlying text is deleted, not just hidden. This is the standard mode for publishing on transparency portals or official gazettes.

Asterisk masking

The value is replaced with asterisks or dots, fully or partially (***4567X for a national ID). Useful when regulations require showing the minimum identifying digits of a contractor or grant recipient.

Suppression or labelling

The value is deleted entirely or replaced with a category tag ([NAME], [ADDRESS], [IBAN]). Keeps the document structure and makes later analysis easy without real data.

Pseudonymize (still usable)

The document stays coherent and exploitable, but nobody is identifiable without the additional information, which is kept separately.

Synthetic data

Every real value is swapped for a fictitious yet valid, coherent one (names, national IDs, addresses or IBANs with correct formats). Ideal for testing, training or sending the document to a generative AI.

Coded pseudonyms

Each person becomes a stable code (Applicant 1, Resident A) across the document or the whole file, so the narrative still holds without identifying anyone.

Reversible tokenization

Data is swapped for tokens while the original stays encrypted in your key vault. Only authorized staff can reverse it, with an auditable log of every re-identification.

Who it's for

Anonymize documents in any sector.

More than 200 public administrations already trust oGov.

From governments to private firms: any organization handling personal data can reduce risk and stay compliant.

Public sector

City councils, provincial governments, universities and agencies that publish minutes, resolutions and files with full legal certainty.

Law and consulting firms

Law firms, consultancies and advisors preparing judicial, contractual or tax documentation before sharing it.

Healthcare

Hospitals, clinics and insurers anonymizing clinical records, reports and studies for research or exchange.

Anyone using AI

Companies training models or using AI assistants that need to strip personal data before processing.

Use cases

For everything that goes public.

Transparency portal

Publish minutes, resolutions, agreements and contracts on the portal without exposing third parties' personal data.

Open data

Anonymize tabular and document datasets before publishing them on your open-data hub.

File access requests

Deliver files to transparency requesters or interested parties with third-party data automatically redacted.

Sending data to AI

Before sending documents to a generative AI model, strip personal data so no external provider processes it.

Compliance

Compliance that backs oGov.

oGov at ENS MEDIUM, servers at ENS HIGH, plus a data governance certificate.

ENS MEDIUM at oGov and ENS HIGH on servers

oGov is certified at MEDIUM level of the Spanish National Security Framework (ENS) and the servers hosting the service are certified at HIGH level, and they also issue a data governance certificate for public administrations.

Auditable, encrypted and traceable

Documents never leave your server. AES-256 encryption, TLS 1.3, processing inside your infrastructure, and full logging of every action with hashes and timestamp.

Free trial

Anonymize your documents in seconds

Upload one and see how personal data is redacted instantly. No sign-up, no card.

Frequently asked questions

Everything about how to anonymize documents

What is document anonymization?

It is the process of irreversibly removing or redacting personal data (names, IDs, addresses, phone numbers, IBAN, license plates) from a document so that individuals cannot be identified. It is a GDPR requirement before publishing administrative files, resolutions or notifications.

What is the difference between anonymization and pseudonymization?

Anonymization is irreversible: personal data disappears and can no longer be linked to an identifiable person, so the document is no longer subject to GDPR. Pseudonymization replaces data with identifiers (e.g. a code), but the link can be reversed with additional information, so it remains personal data processing.

Which document formats does anonimizia support?

With anonimizia you can anonymize PDF (.pdf), Microsoft Word (.docx) and OpenDocument Text (.odt). It returns the document with personal data redacted on the text layer and metadata cleaned, ready to publish on transparency portals or official gazettes.

Is anonimizia GDPR and ENS compliant?

Yes. anonimizia is designed for the public sector and complies with GDPR, LOPDGDD and ENS. Processing runs in your environment, guaranteeing data sovereignty: no document leaves your infrastructure.

Can I anonymize documents before sending them to an AI like ChatGPT?

Yes. Anonymizing first with anonimizia is the recommended pre-processing layer for removing personal data from files, contracts or emails before sending them to generative AI models. You benefit from ChatGPT and other models while complying with GDPR and without exposing sensitive information to third parties.