Babel

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

High-impactSource: a) High-impact

Impact justification

Not reported

Start date

2023-08-29 (day precision)Source: 2023-08-29T00:00:00

Withheld from public reporting

Not reported

Mission

Topic area

Law Enforcement

Operational functions

Translation and transcription, Image and video analysis Derived by keyword rules; see methodology

Problem the AI is intended to solve

Babel utilizes AI modules for text detection and translation as well as object and image recognition to provide analysts with possible matches to manually review in a single interface, versus doing multiple manual queries. The output is not singly used for action or decision making.

Expected benefits

CBP uses this tool to conduct targeted queries to aid CBP in open source research to monitor potential threats or dangers or identify travelers who may be subject to further inspection for violation of laws CBP is authorized to enforce or administer.

System outputs

Babel utilizes AI modules for Text detection and translation as well as object and image recognition to provide analysts with possible matches to manually review in a single interface versus doing multiple manual queries. The output is not singly used for action or decision making and are used to identify additional Open Source or Social Media of a person or identify additional selectors (such as phone and emails) that are previously unknown to CBP and compared by an analyst against Government systems to identify additional derogatory information. These factors can often eliminate the traveler from additional screening.

Technology

AI classification

Natural language processingSource: Natural Language Processing: AI that processes, interprets, and shares information in human language.

System name(s)

Babel Street

Custom-developed code

No

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)

Babel Street

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

www.dhs.gov ↗

Authorization to Operate

Yes

Demographic features

Not reported

Training and evaluation data

Babel uses proprietary data, public datasets, and machine-labeled datasets to train its NLP and matching models. Evaluation data includes human-annotated datasets, precision, recall, and F1 score assessments, and customer-provided labeled name pairs for tuning. All returned results are carefully reviewed, and no sensitive CBP-owned data is involved in the process

Federal Data Catalog

Not reported

Governance

8 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

Potential for inaccurate translation or translation missing context due to local variant dialects or slang that are not captured by the vendor which can be mitigated through coordination with CBP employees who are from that country/area, foreign language certified, and can be consulted for clarification.