U.S. flag

An official website of the United States government, Department of Justice.

Chat Analysis Triage Tool: Enhancing Language-Based and Biometric Analyses for Real-Time Data Involving Child Sexual Exploitation

Award Information

Award #
15PJDP-21-GK-03269-MECP
Location
Awardee County
Tippecano
Congressional District
Status
Open
Funding First Awarded
2021
Total funding (to date)
$850,000

Description of original award (Fiscal Year 2021, $850,000)

Online child sexual exploitation is a global problem that was exacerbated by the COVID-19 pandemic as nearly all social, educational, and professional lives moved online. In 2020, 21.7 million reports were made to NCMEC’s CyberTipline - a 28% increase from 2019. Worldwide, law enforcement specifically experienced an increase in the number of cases involving online enticement and live-stream child sex abuse. These online offenses all involve chats between minors and offenders, and in these chats, the offenders engage in child sexual grooming.   Child sexual grooming is the process by which an adult uses deception and manipulation to lure a child into engaging in sexual behaviors. It is estimated that 1/3 of online child solicitors are contact-driven (motivated to have sex with the minor in the real world). Thus, ICAC investigators need a way to quickly differentiate high-priority contact-driven offenders from fantasy-driven, and need a way to uniquely identify the offender within the solicitation (e.g., biometric features) – all in real-time.   The overall goal of this proposal is to assist law enforcement investigating online enticement and live-stream child sexual abuse cases. This will be achieved by augmenting and enhancing the language-based and biometric capabilities of the Chat Analysis Triage Tool (CATT) for real-time data. CATT is a forensically sound investigative tool that analyzes chats (between a minor and offender) and assists law enforcement in prioritizing cases in which the offender is more likely to be contact-driven. In addition, CATT analyzes the offenders’ images using facial recognition and knuckle biometrics.

Date Created: November 9, 2021