With AI-powered Bulk Redaction, you can efficiently review and redact various types of sensitive information across entire documents in a streamlined table view. It enables teams to move beyond individual document review, surface all potential redactions in a centralized view, and apply consistent standards across hundreds of files.
With In today’s fast-paced digital environment, maintaining the confidentiality and integrity of your documents is more important than ever. This article walks you through each step of the process, from flagging sensitive content to applying redaction and exporting clean versions, helping your organization stay secure and compliant.
ℹ️ The redaction workflow supports PDF and Word documents. OCR-processed PDFs are not currently supported.
How the redaction workflow works
This table acts as the single source of truth for telling the platform what will be redacted.
The table contains two columns:
Redact
Lists the exact text or data that will be redacted from the documents.Reason
Explains why the data is being redacted. This is a select field with pre-populated reasons, and you can also add custom reasons to match your internal or regulatory requirements.
Ways to Add Redactions
You can populate the redaction table in three main ways:
Categories (non‑AI detection)
Dragon AI (AI‑powered detection)
Manual entry
You can mix and match these methods within the same tab.
Categories (Non-AI)
Categories allow you to automatically detect and redact predefined types of sensitive data without using AI. After uploading your files and initiating the matter, select Categories to choose which data types you want to redact.
You can redact either:
All types, or
Specific data types within a category
Available Categories
Personally Identifiable Information (PII)
Address
Age
Email address
Gender
Phone number
US Social Security number
UK National Insurance number
US ZIP code
Financial and Business Information
Account number
Amounts
Credit card number
IBAN
SWIFT / BIC code
Other Sensitive Data
Date
Time
URLs
IP addresses
Eight‑digit numbers
Six‑digit numbers
Once selected, all matching data will automatically appear in the redaction table, ready for review.
Dragon AI
Use Ask AI to redact anything to describe, in plain language, what you want removed. For example:
“Redact all names of individuals except company directors.”
“Redact all bank details and transaction references.”
You’ll also see various Suggestions you can choose from:
Redact all personally identifiable information (PII)
Redact to comply with GDPR requirements
Redact patient health information (HIPAA)
Redact all business sensitive content
When you select a suggestion, the AI highlights the content it has identified.
You can then choose one of the following actions:
Insert – Accept the suggested redactions and add them to the table.
Regenerate – Ask the AI to try again with a revised result.
Discard – Ignore the suggestion entirely.
Reviewing Redactions in the Preview
As you review documents, the preview pane shows a transparent view by default:
Redacted content is highlighted in green, allowing you to clearly see what has been redacted and where it appears in the document.
You can toggle the preview to view the final redacted version, where redactions appear as solid black bars.
This dual‑view approach helps reviewers validate accuracy before saving.
Saving and Exporting Documents
When you save your matter, documents are exported as PDFs with the following options:
Redacted
Final PDFs with full black‑line redactions applied and all metadata is scrubbed.Transparent
PDFs with green highlights, useful for sharing internally to show what was redacted and where.Excel Copy
An audit report exported as an Excel file, containing a breakdown of:What content was redacted
The reason for each redaction
This Excel export supports compliance and audit requirements and reflects the redaction table exactly—an important capability for teams that need to maintain a formal redaction log.
File Naming After Redaction
Exported file names are automatically updated to include “Redacted”.
If any words in the original file name are themselves redacted, those words are removed.
Example:
Project Falcon – Shareholder Resolution.docx
becomes
Redacted – Project Falcon – – Shareholder Resolution.pdf
Best Practices
Review the redaction table carefully before saving to ensure accuracy and consistency.
Use standardized Reasons to maintain a clean audit trail.
Share Transparent versions internally for review, and Redacted versions externally.
Export and retain the Excel copy for compliance, legal review, or regulatory audits.