AI Pilot Project to Screen and Flag for Personally Identifiable Information (PII) in Digitized Archival Records

Overview

Development stage

PilotSource: b) Pilot – The use case has been deployed in a limited test or pilot capacity.

High-impact designation

Not high-impactSource: c) Not high-impact

Impact justification

Not reported

Start date

2024-07-01 (day precision)Source: 2024-07-01T00:00:00

Withheld from public reporting

Yes (disclosure risk)Source: b) Yes – agency has determined that there’s a risk to disclosure, such as a harm to an interest protected by a FOIA exemption

Mission

Topic area

Other

Operational functions

FOIA and records management Derived by keyword rules; see methodology

Problem the AI is intended to solve

The AI is intended to solve the problem of manual processing bottlenecks and privacy risks by automating the identification and redaction of sensitive personal information within massive digitized collections to ensure safe and rapid public access.

Expected benefits

NARA is using AI to automatically find and remove sensitive personal information from its digitized historical records. This will make the process faster, more accurate, and better protect privacy while allowing for greater public access to these records. The AI will also help NARA manage its vast collection and adapt to future privacy needs.

System outputs

This AI use case will produce redacted records, a prioritized list for redaction, and a tool for staff to redact unpublished records.

Technology

AI classification

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

System name(s)

Not reported

Custom-developed code

Yes

Public source code

Not reported

Sourcing

How it was built

Developed in-houseSource: b) Developed in-house

Vendor (as reported)

Not reported

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

No

Privacy Impact Assessment

Not reported

Authorization to Operate

No

Demographic features

k) None of the above

Training and evaluation data

Agency is strictly using its data to test the pre-trained model for the purpose of prototype work. The data is publicly available at catalog.archives.gov.

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