# Records release · Use cases · Dscvry

> Records release: freedom of information and disclosure requests read page by page, with personal and sensitive details found, a reason proposed for each...

Source: https://dscvryai.com/use-cases/records-release/

# What must be redacted before release?

Every page of a release read for names, addresses and sensitive details, with each proposed redaction shown and its reason given, before anything goes out.

See it worked through ·

1. [Home](https://dscvryai.com/)
2. [Use cases](https://dscvryai.com/#use-cases)
3. Records release

## Hundreds of pages. One missed name.

A disclosure can run to hundreds of pages of emails, minutes and scans. One unredacted name or phone number can cause real harm.

### Read line by line.

Officers check every page by hand, often against a tight legal deadline.

### Details hide everywhere.

Names sit in email signatures, handwritten notes, attachments and scanned forms.

### Reasons applied unevenly.

Two officers can redact the same detail differently, or give different reasons for it.

## What the AI reads.

Every page in the release, read for details that identify a person or should not be published.

### In

- The request
- Emails and attachments
- Minutes and reports
- Scanned letters and forms
- Your redaction guidance
- Earlier releases

### At work

1. [Sense](https://dscvryai.com/technology/sense/) Reads typed pages, scans and handwriting across the whole release.
2. [Understand](https://dscvryai.com/technology/understand/) Finds names, addresses, phone numbers and other personal or sensitive details.
3. [Check](https://dscvryai.com/technology/check/) Proposes each redaction with its reason, and checks it matches earlier releases.
4. [Operate](https://dscvryai.com/technology/operate/) Prepares the redacted copy for approval and keeps the original under access control.

### Out

A release with every proposed redaction marked and its reason given.

## Every page read. Every redaction reasoned.

The disclosure officer sees each proposed redaction, why it was proposed, and the page it sits on.

Request FOI-2291 · The reasoning ·

Request FOI-2291 · Depot closure correspondence · 214 pages Illustrative

6 findings · 4 to redact · 1 to keep · 1 for review

| Page | Detail found | Proposal | Detail | Reason |
|---|---|---|---|---|
| p. 12 | Resident’s name and home address | Redact | Personal data of a member of the public | Personal data |
| p. 12 | Name of a senior official | Keep | Named in their public role, as in earlier releases | Public role |
| p. 37 | Mobile number in an email signature | Redact | Direct number of a junior officer | Personal data |
| p. 88 | Handwritten note naming a staff member’s illness | Redact | Health details | Sensitive data |
| p. 141 | Tender prices from a contractor | For review | May be commercially sensitive; needs a decision | Commercial |
| p. 196 | Bank details on a scanned invoice | Redact | Account number and sort code | Personal data |

Illustrative example. Department, request, people and documents are fictional.

## Built for disclosure teams.

The AI finds and proposes. Your officers decide what is released.

### Nothing released automatically.

Every redaction is approved by a person before the release goes out.

How Operate works
https://dscvryai.com/technology/operate/

### Every redaction has a reason.

The detail, the page and the reason sit together, so the decision can be explained later.

How Check works
https://dscvryai.com/technology/check/

### Originals stay inside.

Unredacted records stay under access control in your environment, including air-gapped.

Deployment options
https://dscvryai.com/#deployment

## Same records. Other questions.

The documents and data behind this work answer other questions too.

What personal data is in these files?

### Sensitive data

Find it and remove it before anything is shared.

See how →
https://dscvryai.com/use-cases/sensitive-data/

What are the responses actually saying?

### Consultation responses

Read every response, group the arguments, and quote the evidence for each.

See how →
https://dscvryai.com/use-cases/consultation-responses/

Which cases can be decided today?

### Casework backlog

Clear the routine cases fast, and send the hard ones to a caseworker.

See how →
https://dscvryai.com/use-cases/casework-backlog/

Where is the evidence for the audit?

### Audit evidence

Pull the proof together from every system.

See how →
https://dscvryai.com/use-cases/audit-evidence/

More in Government
https://dscvryai.com/#use-cases

Every use case
https://dscvryai.com/use-cases/

## Bring us a past release and its unredacted pages.

We’ll show you what we would have proposed to redact, and the reason for each.

[Talk to us →](https://dscvryai.com/contact/?topic=Records%20release%20inquiry)

**London**Discovery AI Limited · United Kingdom

**Hyderabad**ReDscvry Technology Private Limited · India

**Chennai**ReDscvry Technology Private Limited · India
