FOIA REDACTION (FRED) TOOL

Overview

Development stage

DeployedSource: c) Deployed – The use case is being actively authorized or utilized to support the functions or mission of an agency.

High-impact designation

Presumed high-impact, determined notSource: b) Presumed high-impact, but determined not high impact

Impact justification

FRED should not be considered a high-impact AI use case because: a. It operates as a recommendation system only, with mandatory human review and approval required for all outputs. b. No automated decisions are made - humans retain full control over all redaction decisions. c. It serves as an assistive tool to improve efficiency while maintaining human oversight and accountability for all FOIA compliance decisions.

Start date

2025-05 (month precision)Source: May-25

Withheld from public reporting

NoSource: a) No

Mission

Topic area

Health and Medical

Operational functions

FOIA and records management Derived by keyword rules; see methodology

Problem the AI is intended to solve

FRED is designed to support FOIA staff in redacting records more efficiently and consistently. FOIA redaction can be a time-consuming process and experience large back-logs of requested documents. It aims to use AI to analyze, identify, and generate predictions of text for redaction, thereby improving the efficiency of FOIA response processing.

Expected benefits

More efficient releases of documents in response to FOIA requests, reduction in back-logs, helping ensure public access to government information while protecting sensitive data appropriately.

System outputs

FRED produces a PDF with boxes around text that it recommends for redaction along with comments for the redaction code

Technology

AI classification

Generative AISource: Generative AI: AI that generates new or synthetic content (e.g., images, videos, audio, text, code).

System name(s)

Elsa

Custom-developed code

Yes

Public source code

Not reported

Sourcing

How it was built

Contract and in-houseSource: c) Developed with both contracting and in-house resources

Vendor (as reported)

Contractor teams

Vendors (standardized)

None on the standard list

A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.

Data and privacy

Involves PII

Yes

Privacy Impact Assessment

Not reported

Authorization to Operate

Yes

Demographic features

k) None of the above

Training and evaluation data

Completed 483 forms before redaction and versions of those forms after redaction by FDA staff

Federal Data Catalog

Not reported

Governance

0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.

Potential impacts and how they were identified

Not reported