By · Last updated 2026-05-29

Back to BlogTechnical

Cross-Application PII: Word, Chrome, and AI

Customer data flows from browser research to Word drafts to Claude prompts. Each context switch is a potential leakage point.

May 29, 20266 minute read
cross-platform PIIOffice Add-inChrome extensionMCP Serverworkflow privacy

Cross-Application PII: Word, Chrome, and AI

Updated for 2026.

Customer data doesn't stay in one place. It moves between apps as part of normal work. Each move is a chance for that data to leak.

The Multi-App Data Flow Problem

Think about how a legal researcher works. They look up case details in Chrome. They copy those details into Word. Then they paste excerpts into Claude for help with a draft. Client names travel from app to app at each step.

A support manager does the same. They open a customer complaint in a browser CRM. They copy it into Word for internal escalation. Then they paste it into an AI tool to draft a reply. The customer's name and account details move through three apps.

An HR professional downloads employee records to Excel. They open the file and run some analysis. Then they paste summaries into PowerPoint for a leadership meeting. Employee PII lives in every app along the way.

All of these workflows share one trait. The same PII exists in several places at once. Each app switch is a new chance for exposure — in an AI prompt, a screenshot, an email attachment, or a shared file.

Why One-App Protection Fails

A Chrome extension that guards AI prompts is useful. But it only works in the browser. The same customer data it blocks from ChatGPT can still:

  • Appear in a Word file sent to outside counsel
  • Get pasted into Teams chat with no warning
  • End up in an Excel file in a shared cloud folder

An Office Add-in that guards Word is useful. But it only works in Word. Client names in that document can still be pasted into Claude Desktop. No detection runs. No warning appears.

One tool covering one app leaves every other app exposed. PII leaks through the gaps.

Where Protection Is Needed

Start by mapping PII flows across all apps your team uses.

Common flows to map:

  • Browser (CRM or portal) → Word (reports or letters)
  • Browser (research) → AI tool (drafting or summary)
  • Email → Word (complaint documentation)
  • Excel (exported data) → AI tool (analysis)
  • Word or PDF → AI tool (review or drafting)
  • Any app → Screenshot → Collaboration tool

For each flow, ask: where does protection apply, and where are the gaps?

Protection by tool:

  • Browser AI prompt: Chrome Extension
  • Word and Excel: Office Add-in
  • Claude Desktop or Cursor: MCP Server
  • Bulk file processing: Desktop App or Web App
  • Images and screenshots: Image PII detection

Any flow that passes through an unprotected step has a gap. That gap needs to be closed.

Use the Same Detection Engine Everywhere

Cross-app protection only works if the same engine runs in every context.

If the Chrome Extension uses a different engine than the Office Add-in, problems arise. The same name may be caught in Chrome but missed in Word. Confidence scores may differ. Replacement tokens may differ too. That makes it impossible to trace data across documents.

Good cross-app protection uses the same model, the same entity types, the same thresholds, and the same replacement logic — in every app.

A legal researcher uses three tools each day:

  • Microsoft Word for drafting opinions
  • Chrome for looking up case law via Claude
  • Claude Desktop for AI-assisted drafting

Client names and case references flow through all three tools in a normal day.

Before setup:

  • Chrome Extension installed: AI prompts in Chrome are protected
  • No Office Add-in: client names in Word are not protected when shared
  • No MCP Server: client names in Claude Desktop are not protected

After setup with a shared preset:

  • Chrome Extension: catches client names before AI submission
  • Office Add-in: catches client names before email or external sharing
  • MCP Server: catches client names before Claude Desktop receives them

The key: One "Legal Research" preset — set up once — runs the same way in all three apps. A name caught in Word is caught the same way in Chrome and in Claude Desktop.

When the preset is updated, the change flows to all three apps through shared configuration. There is nothing to maintain separately.

For more on preset-based detection, see how anonymization presets work across GDPR audit contexts.

Start with the Highest-Risk Flows

Not all flows carry the same risk. Start where exposure is highest.

Tier 1 — protect first:

  • AI tool flows (PII leaves your controlled systems)
  • External sharing flows (email attachments, cloud storage links)
  • Regulatory reporting flows (data sent to authorities or third parties)

Tier 2 — protect next:

  • Internal collaboration flows (documents seen by many team members)
  • Data export flows (database exports, system reports)

Tier 3 — lower urgency:

  • Internal file creation (documents not shared outside)
  • Local analysis (Excel work for internal reporting only)

Tier 1 has the most exposure under GDPR Article 32. It also gives the most risk reduction per unit of effort.

For a full look at GDPR Article 32 requirements, see GDPR compliance technical controls.

To see how multi-surface protection works in practice, see cross-platform PII compliance across Mac, Linux, and Windows.

Sources

Ready to protect your data?

Start anonymizing PII with 285+ entity types across 48 languages.

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.

Related reading

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.