Guides

Practical guidance on using AI safely with confidential documents — anonymising, pseudonymising, and what GDPR has to say about it.

• How to anonymise a document for ChatGPT (and keep it useful)

  Want to use ChatGPT on a confidential document? Anonymise or pseudonymise it first. What to shield, how to do it in one minute, and where manual redaction fails.

• Can I upload client data to ChatGPT? What GDPR actually says

  Putting client data into ChatGPT is GDPR-regulated processing, even on a business plan. When it is allowed, when it is not, and how to make a document safe first.

• How to remove personal data from a document: 4 methods compared

  Black bars, manual rewriting, Word's inspector or automatic anonymisation. What actually works to get personal data out of a PDF or Word file.

• Pseudonymisation vs anonymisation: the difference, and when to use which

  Pseudonymisation is reversible with a key, anonymisation is not. What GDPR says about each, and which to pick for AI use, sharing or publication.

• Is it safe to upload confidential documents to AI?

  What actually happens when you upload a confidential document to ChatGPT, Claude or Gemini, where the real risks sit, and the workflow that removes most of them.

• A GDPR-safe AI document workflow for consultants, in 5 steps

  How advisory firms use ChatGPT and Claude on client documents without exposing client data. A 5-step workflow you can write into your processing register.

• ChatGPT Team and confidential documents: what it solves and what it doesn't

  ChatGPT Team doesn't train on your data. Good. But you still send raw client data. What a business plan does and doesn't cover, and what belongs next to it.